PAINTING

FIELD .AND BACKGROUND OF THE INVENTION

The present invention relates in general to a method for simultaneous detection of multiple fluorophores. More particularly, the present invention relates to a spectral imaging method aimed at detecting and analyzing fluorescent in situ hybridizations employing numerous chromosome paints and/or loci specific probes each labeled with a different fluorophore or a combination of fluorophores, the method is highly sensitive both in spatial and spectral resolutions and is capable of simultaneous detection of dozens of flourophores and/or combinations of flourophores, therefore, the method of the present invention can be used for the detection of fluorescently painted complete sets of chromosomes and/or multiple loci from a species such as human and to provide a complete color karyotype.

A spectrometer is an apparatus designed to accept light, to separate (disperse) it into its component wavelengths, and measure the lights spectrum, that is the intensity of the light as a function of its wavelength. An imaging spectrometer is one which collects incident light from a scene and measures the spectra of each pixel (i.e., picture element) thereof.

Spectroscopy is a well known analytical tool which has been used for decades in science and industry to characterize materials and processes based on the spectral signatures of chemical constituents. The physical basis of spectroscopy is the interaction of light with matter. Traditionally, spectroscopy is the measurement of the light intensity emitted, transmitted, scattered or reflected from a sample, as a function of wavelength, at high spectral resolution, but without any spatial information.

Spectral imaging, on the other hand, which is a combination of high resolution spectroscopy and high resolution imaging (i.e., spatial information) has yet not been used for analyzing biological samples. The closest work so far described concerns either obtaining high spatial resolution information from a biological sample yet providing only limited spectral information, for example, when high spatial resolution imaging is performed with one or several discrete band-pass filters [See, Andersson-Engels et al. (1990) Proceedings of SPIE - Bioimaging and Two-Dimensional Spectroscopy, 1205, pp. 179-189], or alternatively, obtaining high spectral resolution (e.g., a full spectrum), yet limited in spatial resolution to a small number of points of the sample or averaged over the whole sample [See for example, U.S. Pat. No. 4,930,516, to Alfano et al.].

As will be described in great details below, combining spectroscopy with imaging is useful for various biological research and medical applications and is referred to hereinbelow as spectral bio-imaging. One example for the usefulness of spectral bio-imaging concerns detection of specific cellular constituents (e.g., proteins, nucleic acid sequences, etc.) after being labeled (i.e., tagged) with fluorescent probes. In this direction spectral imaging can be used to identify and map several fluorophores simultaneously in one measurement. In fact, the inherently high spectral resolution of spectral imaging of the present invention is ideally suited for 'sorting out' fluorescent probes (or other chemical constituents) with overlapping spectra.

Conceptually, a spectral bio-imaging system consists of (1) a measurement system, and (2) an analysis software. The measurement system includes all of the optics, electronics and the manner in which the sample is illuminated (e.g., light source selection), the mode of measurement (e.g., fluorescence), as well as the calibration best suited for extracting the desired results from the measurement. The analysis software includes all of the software and mathematical algorithms necessary to analyze and display important results in a meaningful way.

Spectral imaging has been used for decades in the area of remote sensing to provide important insights in the study of Earth and other planets by identifying characteristic spectral absorption features. However, the high cost, size and configuration of remote sensing spectral imaging systems (e.g., Landsat, AVTRJS) has limited their use to air and satellite-borne applications [See, Maymon and Neeck ( 1988) Proceedings of SPIE - Recent Advances in Sensors. Radiometry and Data Processing for Remote Sensing, 924, pp. 10-22; Dozier ( 1988) Proceedings of SPIE - Recent Advances in Sensors, Radiometry and Data Processmg for Remote Sensmg, 924, pp. 23-30]

There are three basic types of spectral dispersion methods that might be considered for a spectral bio-imaging system: (i) spectral grating, (ii) spectral filters and (iϋ) interferometric spectroscopy. As will be described below, the latter is best suited to implement the method of the present invention.

In a grating (i.e., monochromator) based systems, also known as slit-type imaging spectrometers, such as for example the DILOR system: [see, Valisa et al. (Sep. 1995) presentation at the SPIE Conference European Medical Optics Week, BiOS Europe '95, Barcelona, Spain], only one axis of a CCD (charge coupled device) array detector (the spatial axis) provides real imagery data, while a second (spectral) axis is used for sampling the intensity of the light which is dispersed by the grating as function of wavelength. The system also has a slit in a first focal plane, limiting the field of view at any given time to a line of pixels. Therefore, a

full image can only be obtained after scanning the grating or the incoming beam in a direction parallel to the spectral axis of the CCD in a method known in the literature as line scanning. The inability to visualize the two-dimensional image before the whole measurement is completed makes it impossible to choose, prior to making a measurement, a desired region of interest from within the field of view and/or to optimize the system focus, exposure time, etc. Grating based spectral imagers are in use for remote sensing applications, because an airplane (or satellite) flying over the surface of the Earth provides the system with a natural line scanning mechanism. It should be further noted that slit-type imaging spectrometers have a major disadvantage since most of the pixels of one frame are not measured at any given time, even though the fore-optics of the instrument actually collects incident light from all of them simultaneously. The result is that either a relatively large measurement time is required to obtain the necessary information with a given signal-to-noise ratio, or the signal-to-noise ratio (sensitivity) is substantially reduced for a given measurement time. Furthermore, slit-type spectral imagers require line scanning to collect the necessary information for the whole scene, which may introduce inaccuracies to the results thus obtained.

Filter based spectral dispersion methods can be further categorized into discrete filters and tunable filters. In these types of imaging spectrometers the spectral image is built by filtering the radiation for all the pixels of the scene simultaneously at a different wavelength at a time by inserting in succession narrow band filters in the optical path, or by electronically scanning the bands using acousto-optic tunable filters (AOTF) or liquid-crystal tunable filter (LCTF), see below. Similarly to the slit type imaging spectrometers equipped with a grating as described above, while using filter based spectral dispersion methods, most of the radiation is rejected at any given time. In fact, the measurement of the whole image at a specific wavelength is possible because all the photons outside the instantaneous wavelength being measured are rejected and do not reach the CCD. The sensitivity advantage that interferometric spectroscopy has over the filter and grating method is known in the ait as the multiplex or Fellgett advantage.

Tunable filters, such as AOTFs and LCTFs have no moving parts and can be tuned to any particular wavelength in the spectral range of the device in which they are implemented. One advantage of using tunable filters as a dispersion method for spectral imaging is their random wavelength access; i.e., the ability to measure the intensity of an image at a number of wavelengths, in any desired sequence without the use of filter wheels. However, AOTFs and LCTFs have the disadvantages of (i) limited spectral range (typically, λ _max = 2λ _πml ) while all

other radiation that falls outside of this spectral range must be blocked, (u) temperature sensitivity, (in) poor transmission, (ι\) polarization sensitivity, and (v) m the case of AOTFs an effect of shifting the image duπng wavelength scanning

All these types of filter and tunable filter based systems have not been used successfully and extensively over the years m spectral imaging for any application, because of their limitations in spectral resolution, low sensitivity, and lack of easy- to-use and sophisticated software algorithms for interpretation and display of the data

A method and apparatus for spectral analysis of images which have advantages m the above respects was disclosed m U S Pat application No 08/392,019 to Cabib et al , filed Feb 21, 1995, now U S Pat No 5,539,517, issued Jul 23, 1996, which is incorporated by reference as if fully set forth herein, w th the objective to provide a method and apparatus for spectral analysis of images which better utilizes all the information available from the collected mcident hght of the image to substantially decrease the required frame time and/or to substantially increase the signal-to-noise ratio, as compared to the conventional slit- or filter t pe imaging spectrometer and does not mvolve line scanning Accordmg to this invention, there is piovided a method of analyzing an optical image of a scene to determine the spectral intensity of each prxel thereof by collectmg mcident hght from the scene, passmg the light through an mterferometer which outputs modulated hght correspondmg to a predetermined set of linear combinations of the spectral intensity of the light emitted from each pixel, focusmg the light outputted from the mterferometer on a detector array, scanning the optical path difference (OPD) generated m the mterferometer for all pixels mdependently and simultaneously and processmg the outputs of the detector array (the lnterferograms of all pixels separately) to determine the spectral intensity of each pixel thereof This method may be practiced by utilizing types of interferometers wherem the OPD is varied to build the interferograms by moving the entire mterferometer, an element withm the mterferometer, or the angle of mcidence of the incoming radiation In all of these cases, when the scanner completes one scan of the mterferometer, the interferograms for all pixels of the scene are completed Apparatuses in accordance with the above features differ from the conventional sht- and filter type imaging spectrometers by utilizing an mterferometer as descnbed above, therefore not limiting the collected energy with an aperture or sht or limiting the mcoπung wavelength with narrow band interference or tunable filters, thereby substantially mcreasmg the total throughput of the system Thus, mterferometer based apparatuses better utilize all the information available from the mcident light of the scene to be analyzed, thereby

substantially decreasing the measuring time and/or substantially increasing the signal-to-noise ratio (i.e., sensitivity). Consider, for example, the "whisk broom" design described in John B. Wellman ( 1987) Imaging Spectrometers for Terrestrial and Planetary Remote Sensing, SPIE Proceedings, Vol. 750, p. 140. Let n be the number of detectors in the linear array, m x m the number of pixels in a frame and T the frame time. The total time spent on each pixel in one frame summed over all the detectors of the array is nTlm ^~ . By using the same size array and the same frame rate in a method according to the invention described in U.S. Pat. application No. 08/392,019, the total time spent summed over all the detectors on a

9 particular pixel is the same, nTlrrr However, whereas in the conventional grating method the energy seen by every detector at any time is of the order of \in of the total, because the wavelength resolution is \ln of the range, in a method according to the invention described in U.S. Pat. application No. 08/392,019 the energy is of the order of unity because the modulating function is an oscillating function (e.g., sinusoidal (Michelson) or similar periodic function such as low finesse Airy function with Fabry-Perot) whose average over a large OPD range is 50%. Based on the standard treatment of the Fellgett advantage (or multiplex advantage) described in lnterferometry textbooks [for example, see, Chamberlain ( 1979) The principles of interferometric spectroscopy, John Wiley and Sons, pp. 16- 18 and p. 263], it is possible to show that devices according to this invention have measurement signal-to-noise ratios which are improved by a factor of n in the cases of noise limitations in which the noise level is independent of signal (system or background noise limited situations) and by the square root of the ratio of the signal at a particular wavelength to the average signal in the spectral range, at wavelengths of a narrow peak in the cases the limitation is due to signal photon noise. Thus, according to the invention described in U.S. Pat. application No. 08/392,019, all the required OPDs are scanned simultaneously for all the pixels of the scene in order to obtain all the information required to reconstruct the spectrum, so that the spectral information is collected simultaneously with the imaging information. This invention can be used with many different optical configurations, such as a telescope for remote sensing, a microscope for laboratory analysis, fiber optics for industrial monitoring and medical imagmg, diagnosis, therapy and others.

In a continuation application (U.S. Pat. application 08/571,047, to Cabib et al., filed Dec. 12, 1995 which is incorporated by reference as if fully set forth herein) the objective was to provide spectral imaging methods for biological research, medical diagnostics and therapy, which methods can be used to detect spatial organization (i.e., distribution) and to quantify cellular and tissue natural

constituents, structures, organelles and administered components such as tagging probes (e.g., fluorescent probes) and drugs using Hght transmission, reflection, scattering and fluorescence emission strategies, with high spatial and spectral resolutions. In U.S. Pat. application 08/571,047, the use of the spectral imaging apparatus described in U.S. Pat. application No. 08/392,019 for interphase fluorescent in situ hybridization of as much as six loci specific probes (each loci located on a different chromosome) was demonstrated, as well as additional biological and medical applications.

Spectral bio-imaging systems are potentially useful in all applications in which subtle spectral differences exist between chemical constiments whose spatial distribution and organization within an image are of interest. The measurement can be carried out using virtually any optical system attached to the system described in patent application No. 08/392,019, for example, a fluorescence microscope combined with administered fluorescent fluorophores or combinations of fluorophores.

Fluorescence measurements can be made with any standard filter cube (consisting of a barrier filter, excitation filter and a dichroic mirror), or any customized filter cube or combinations of filter cubes for special applications, provided the emission spectra fall within the spectral range of the system sensitivity.

One of the major benefits of the Human Genome Project (HGP) has been the isolation of a large number of nucleic acid probes for diseased genes and other chromosome regions and structures. This has stimulated interest in DNA diagnostics as the number and types of tests that can be developed is dependent upon these probes. In recent years there has been particular interest in fluorescent in situ hybridization (FISH) which is the process of marking with a fluorescent moiety conjugated to a specific nucleic acid molecule complementary to an examined chromosome region (collectively referred herein as a probe), followed by visualization of the fluorescent moiety by fluorescence microscopy. There is a clear trend for employing FISH technology in the clinic in parallel to its traditional employment in the basic research laboratory. FISH may be considered an advanced approach to cytogenetics and it is clear that the amount of information about chromosomes that may be gained from FISH far outdistances that obtained from standard karyotyping by DNA banding methods. In addition, diagnostics information may be gained much more rapidly using techniques such as interphase cytogenetics as compared to classical (metaphase) cytogenetics.

According to the present invention provided is a FISH imaging method, capable of simultaneously acquire fluorescence spectra from all pixels of a field of

view of a fluorescence microscope and simultaneously detect the location of dozens of probes in a single measurement. In conjunction with the availability of chromosome specific probes (i.e., chromosome paints) and novel labeling strategies, the method is able to create a FISH karyotype with each chromosome being painted with a different color (i.e., 24 different colors for a human male karyotype, 23 for a female). This method results in extremely h gh sample throughput and allows analysis of an essentially unlimited number of probes.

There is thus a widely recognized need for, and it would be highly advantageous to have a spectral imaging method for detecting and analyzing fluorescent in situ hybridizations employing numerous chromosome paints and/or loci specific probes each labeled with a different fluorophore or a combination of fluorophores for the detection of fluorescently painted complete sets of chromosomes and/ or multiple loci from a species such as human.

SUMMARY OF THE INVENTION

According to the present invention there is provided a spectral imaging method aimed at detecting and analyzing fluorescent in situ hybridizations employing numerous chromosome paints and/or loci specific probes each labeled with a different fluorophore or a combination of fluorophores, the method is highly sensitive both in spatial and spectral resolutions and is capable of simultaneous detection of dozens of flourophores and/or combinations of flourophores and thus can be used for the detection of fluorescently painted complete sets of chromosomes and/or multiple loci from a species such as human and to provide a color karyotype. According to further features in preferred embodiments of the invention described below, the method comprising the steps of (a) preparing a sample to be spectrally imaged; (b) viewing the sample through an optical device, the optical device being optically connected to an imaging spectrometer, the optical device and the imaging spectrometer being for obtaining a spectrum of each pixel of the sample by (i) collecting incident light simultaneously from all pixels of the sample using collimating optics; (ii) passing the incident collimated hght through an interferometer system having a number of elements, so that the hght is first split into two coherent beams which travel in different directions inside the interferometer and then the two coherent beams recombine to interfere with each other to form an exiting hght beam; (iii) passing the exiting hght beam through a focusing optical system which focuses the exiting hght beam on a detector having a two-dimensional array of detector elements, so that at each instant each of the detector elements is the image of one and always the same pixel of the sample for

the entire duration of the measurement, so that the real image of the sample is stationary on the plane of the detector array and at any time during the measurement the image is still visible and recognizable, and so that each of the detector elements produces a signal which is a particular lmear combmation of hght intensity emitted by the pixel at different wavelengths, wherem the linear combmation is a function of the instantaneous optical path difference, (iv) rotating or translating (1 e , scanning) one or more of the elements of the mterferometer system, so that the optical path difference between the two coherent beams generated by the mterferometer system is scanned simultaneously for all the pixels of the sample, and (v) recordmg signals of each of the detector elements as function of time usmg a recordmg device to form a first spectral cube of data, and (c) interpreting the first spectral cube of data usmg a mathematical algorithm

Accordmg to still further features m the descπbed preferred embodiments the method further comprising the step of (d) displaying a map of the interpreted spectral cube of data

Accordmg to still further features in the described preferred embodiments the optical device is a fluorescence microscope

Accordmg to still further features m the descnbed preferred embodiments the collimated light is a fluorescence light emitted from the sample Accordmg to still f rther features m the descπbed prefeπed embodiments the collimated hght emitted from the sample is an administered probe fluorescence

Accordmg to still further features m the descπbed prefeπed embodiments the hght onginates from a source such as laser, white light, filtered light, ultraviolet light or a light having a small wavelength range

Accordmg to still further features m the descπbed prefeπed embodiments the light oπgmates from a multiplicity of light sources, the sources operate simultaneously or successively

Accordmg to still further features m the descπbed prefeπed embodiments the two-dimensional aπay is selected from the group consistmg of a video rate CCD, a cooled high dynamic range CCD, an intensified CCD and a time gated intensified CCD

Accordmg to still further features m the descπbed prefeπed embodiments the sample is a cell during interphase, a cell duπng mitosis and/or a cell duπng meiosis

Accordmg to still further features m the descπbed prefeπed embodiments the cell is from a human

According to still further features in the described prefeπed embodiments the cell is a cancerous cell, a blood cell, a fetal cell or a cell suspected of being malignant.

According to still further features in the described prefeπed embodiments the sample is a cell, the hght is induced by a probe, the probe binds to a specific cellular constituent, the method is for detecting the presence or the level of the cellular constituent.

According to still further features m the described prefeπed embodiments the probe includes a conjugated fluorescent moiety and the induction is a fluorescence light emission of the fluorescent moiety.

According to still further features in the described prefeπed embodiments the probe further includes a nucleic acid molecule, the method is for detecting the presence or the level of a cellular nucleic acid hybridizing with the nucleic acid molecule. According to still further features in the described prefeπed embodiments the cellular nucleic acid is deoxvribonucleic acid and/or ribonucleic acid.

According to still further features in the described prefeπed embodiments the fluorescent moiety is SpectrumOrange™, SpectrumGreen™, Aqua, Texas-Red, FITC, rhodamine, fluorescein, cascade blue and/or any combination thereof. According to still further features in the described prefeπed embodiments the mathematical algorithm is a point operation analysis of the spectrum of each of the pixels in the sample.

According to still further features m the described prefeπed embodiments the point operation analysis mcludes mapping the spectrum of each of the pixels in the sample into a scalar according to a transformation function.

According to still further features m the described prefeπed embodiments the point operation analysis includes mapping the spectrum of each of the pixels of the sample into another spectrum according to a transformation function.

According to still further features in the described prefeπed embodiments the mathematical algorithm is a morphological analysis.

Accordmg to still further features in the described prefeπed embodiments the mathematical algorithm is a similarity mapping analysis for computing for each of the pixels in the sample a spectral difference from a reference spectrum.

According to still further features in the described prefeπed embodiments the similarity mapping analysis results in generating a gray level or a pseudocolor image, in which bright pixels coπespond to a small spectral difference and dark- pixels coπespond to a large spectral difference.

Accordmg to still further features m the descnbed prefeπed embodiments the similanty mappmg analysis results m generating a gray level or a pseudocolor image, m which bright pixels coπespond to a large spectral difference and dark pixels coπespond to a small spectral difference. Accordmg to still further features m the descπbed prefeπed embodiments the spectral difference is a scalar defined as the integral over a predefined wavelength range of the absolute value of the difference between the spectrum of each of the pixels and the reference spectrum

Accordmg to still further features m the descnbed prefeπed embodiments the mathematical algorithm is a classification mappmg analysis computmg for the spectrum of each of the pixels a spectral difference from several reference spectra.

Accordmg to still further features m the descπbed prefeπed embodiments the classification mappmg analysis results m generating a multicolor image, m which groups of pixels having a predetermined maximal spectral differences from one of the several reference spectra are colored with a predetermined artificial color

Accordmg to still further features m the descπbed prefeπed embodiments the spectral difference is a scalar defined as the integral over a predefined wavelength range of the absolute value of the difference between the spectrum of each of the pixels and one of the several reference spectra

Accordmg to still further featuies m the descπbed prefeπed embodiments the mathematical algonthm is a principal component analysis

Accordmg to still further featuies m the descπbed prefeπed embodiments the principal component analysis mcludes (a) building a covaπant matrix for all of the pixels and the wavelengths of the measurement, mcludmg wavelengths of exciting sources when multiple wavelengths are used, (b) diagonahzing the covanant matrix and finding all independent orthogonal spectral base elements; (c) finding which of the base elements tag certain features m the sample

Accordmg to still further features in the described prefeπed embodiments the mathematical algorithm is a linear combmation analysis

According to still further features m the descπbed prefeπed embodiments the linear combmation analysis mcludes applymg an arithmetical function between coπesponding wavelengths of coπespondmg pairs of pixels belonging to the first spectral cube of data and to a second spectral cube of data, to obtain a resulting third spectral cube of data.

Accordmg to still further featuies m the descnbed prefeπed embodiments the linear combmation analysis is for a purpose such as averaging two spectral cubes of data or tune changes follow-up and spectral normalization

Accordmg to still further features m the descnbed prefeπed embodiments the linear combmation analysis mcludes applymg a given scalar to every wavelength of the spectra of each of the pixels by an anthmetical function, the function is addition, subtraction, multiplication, division and/or combinations thereof

Accordmg to still further features m the descnbed prefeπed embodiments the linear combmation analysis is for background subtraction m which a spectrum of a pixel located m a background region of the sample is subtracted from the spectra of the pixels of the sample Accordmg to still further features m the descnbed prefeπed embodiments the linear combmation analysis is for a calibration procedure m which a spectrum measured pnor to the viewing the sample is for dividing the spectra of the pixels of the sample

Accordmg to still further features in the descπbed prefeπed embodiments the mathematical algonthm is an optical density analysis

Accordmg to still further features m the descnbed prefeπed embodiments the optical density analysis is for obtaining an interpreted image which is an optical density map

Accordmg to still further features m the descπbed prefeπed embodiments the mathematical algonthm computes a Red-Green-Blue color unage usmg predefined wavelength ranges

Accordmg to still further features m the descnbed prefeπed embodiments the Red-Green-Blue color image is modified by a contrast stretching algonthm

Accordmg to still further features m the descπbed prefeπed embodiments the Red-Green-Blue color unage is modified by a contrast stretching algonthm

Accordmg to still further features m the descnbed prefeπed embodiments the mathematical algonthm computes a ratio between mtensities at two different wavelengths for each of the spectra of the pixels

Accordmg to still further features m the descπbed prefeπed embodiments the mathematical algonthm computes a ratio between mtensities at two different wavelengths for each of the spectra of the pixels and pamts each of the pixels in a lighter or darker artificial color, accordmg to the computed ratio

According to still further features m the descπbed prefeπed embodiments the method is for spectral identification of multiple fluorophores administered to the sample

Accordmg to still further features m the descπbed prefeπed embodiments descπbed below the method is a fluorescent in situ hybndization method compπsmg the steps of (a) providing a cell nuclei having chromosomes, the

chromosomes bemg hybndized with at least one nucleic acid probe, each of the at least one nucleic acid probe mcludmg at least one nucleic acid molecule, each of the at least one nucleic acid molecule bemg labeled with at least one fluorophore, (b) viewing the cell nuclei through a fluorescence microscope, the fluorescence microscope bemg optically connected to an imagmg spectrometer, the fluorescence microscope and the imaging spectrometer bemg for obtaining a spectrum of each pixel of the cell nuclei by (1) collecting mcident light simultaneously from all pixels of the cell nuclei usmg collimating optics, (u) passmg the mcident collimated hght through an mterferometer system having a number of elements, so that the light is first split mto two coherent beams which travel in different directions mside the mterferometer and then the two coherent beams recombme to interfere with each other to form an exiting light beam, (m) passmg the exiting light beam through a focusmg optical system which focuses the exiting hght beam on a detector having a two-dimensional aπay of detector elements, so that at each instant each of the detector elements is the image of one and always the same pixel of the cell nuclei for the entire duration of the measurement, so that the real image of the cell nuclei is stationary on the plane of the detector aπay and at any time duπng die measurement the image is still visible and recognizable, and so that each of the detector elements produces a signal which is a particular linear combmation of hght intensity emitted by the pixel at different wavelengths, wherem the linear combmation is a function of the instantaneous optical path difference, (iv) rotating or translating (1 e , scanning) one or more of the elements of the mterferometer system, so that the optical path difference between the two coherent beams generated by the mterferometer system is scanned simultaneously for all the pixels of the cell nuclei, and (v) recordmg signals of each of the detector elements as function of time usmg a recordmg device to form a first spectral cube of data, and (c) interpreting the first spectral cube of data usmg a mathematical algonthm

Accordmg to still further features in the descπbed prefeπed embodiments the at least one nucleic acid molecule is at least one locus, at least one fragmented chromosome, at least one yeast artificial chromosome mcludmg an insert, at least one plasmid mcludmg an insert, at least one cosmid mcludmg an insert, at least one phagemid mcludmg an insert, at least one viral vector mcludmg an insert, a complete genome of a species, a complete genome of a cancerous tissue and/or combinations thereof

Accordmg to still further features m the descnbed prefeπed embodiments the at least one fluorophore is at least one fluorescent combmatonal dye

Accordmg to still further features m the descnbed prefeπed embodiments the cell nuclei is a cell nuclei during interphase, a cell nuclei duπng mitosis and/or a cell nuclei duπng meiosis

Accordmg to still further features m the descπbed prefeπed embodiments the number of nucleic acid probes is one, two, three, four, five, six, seven, eight, nine, ten, eleven, twelve, thirteen, fourteen, fifteen, sixteen, seventeen, eighteen, nineteen, twenty, twenty one. twenty two, twenty three, twenty four or higher than twenty four, each of the probes mcludes a different fluorophore or a different combmation of the fluorophores Accordmg to still further features m the descnbed prefeπed embodiments die chromosomes are interphase chromosomes, chromosomes duπng mitosis and chromosomes duπng meiosis

Accordmg to still further features in the descπbed prefeπed embodiments the mathematical algonthm is a pomt operation analysis of the spectrum of each of the pixels m the cell nuclei

Accordmg to still further features m the descnbed prefeπed embodiments the pomt operation analysis mcludes mappmg the spectrum of each of the pixels m the cell nuclei mto a scalar accordmg to a transformation function

Accordmg to still further features m the descπbed prefeπed embodiments the pomt operation analysis mcludes mappmg the spectrum of each of the pixels of the cell nuclei mto another spectrum accordmg to a transformation function

Accordmg to still further features m the descnbed prefeπed embodiments the mathematical algonthm is a morphological analysis, the morphological analysis determines the relative size of the chromosomes m the cell nuclei Accordmg to still further features m the descnbed prefeπed embodiments the mathematical algonthm is a classification mappmg analysis computmg for the spectrum of each of the pixels a spectral difference from at least one reference spectrum

Accordmg to still further features m the descnbed prefeπed embodiments the classification mappmg analysis results m generating a multicolor unage, m which groups of pixels having a predetermined maximal spectral differences from one of the several reference spectra are colored with a predetermined artificial color

Accordmg to still further features m the descnbed prefeπed embodiments the spectral difference is a scalar defined as the mtegral over a predefined wavelength range of the absolute value of the difference between the spectrum of each of the pixels and one of the several reference spectra

Accordmg to still further features in the descπbed prefeπed embodiments die mathematical algonthm is a principal component analysis

Accordmg to still further features m the descπbed prefeπed embodiments the principal component analysis mcludes (a) building a covanant matrix for all of the pixels and d e wavelengths of the measurement, mcludmg wavelengths of exciting sources when multiple wavelengths are used, (b) diagona zing the covanant matrix and finding all mdependent orthogonal spectral base elements, (c) finding which of the base elements or a combmation thereof tag certain features m the cell nuclei Accordmg to still further features m the descπbed prefeπed embodiments d e mathematical algonthm is a lmear combmation analysis

Accordmg to still further features m the descπbed prefeπed embodiments the lmear combmation analysis is for spectral normalization

Accordmg to still further features m the descnbed prefeπed embodiments the lmear combmation analysis mcludes applymg a given scalar to every wavelength of the spectra of each of the pixels by an aπthmeπcal function, the function is addition, subtraction, multiplication, division and/or combinations thereof

According to still further features m the descπbed prefeπed embodiments d e linear combmation analysis is for background subtraction m which a spectrum of a pixel located m a background region of the cell nuclei is subtracted from the spectra of the pixels of the cell nuclei

Accordmg to still further features m the descnbed prefeπed embodiments me lmear combmation analysis is for a calibration procedure m which a spectrum measured pnor to the viewing the cell nuclei is for dividing the spectra of the pixels of the cell nuclei

Accordmg to still further featuies m the descπbed prefeπed embodiments the mad ematical algonthm computes a Red-Green-Blue color unage usmg predefined wavelength ranges Accordmg to still further features m the descπbed prefeπed embodiments the mathematical algonthm computes a ratio between mtensities at two different wavelengths for each of the spectra of the pixels

Accordmg to still further features in the descnbed prefeπed embodiments die mathematical algonthm computes a ratio between mtensities at two different wavelengths for each of the spectra of the pixels and pamts each of the pixels m a lighter or darker artificial color, accordmg to the computed ratio

Accordmg to still further features m the descπbed prefeπed embodiments the method is for an application such as providing a color karyotype of embryonic

cells, providing a color karyotype of white blood cells, providing a color karyotype of malignant cells and/or providing a color karyotype of cells examined for malignancy.

According to still further features m the described prefeπed embodiments die embryonic cells are chorionic villi cells and/or embryonic cells isolated from a pregnant woman peripheral blood.

According to still further features in the described prefeπed embodiments tihe method is for detecting a trisomy of human chromosome 21. human chromosomal band 21q22, a fragment of human chromosomal band 21q22, human chromosome 18, a fragment of human chromosome 18, human chromosome 13 and a fragment of human chromosome 13.

According to still further features in the described prefeπed embodiments d e providing the color karyotype of the cells examined for malignancy is for obtaining a color translocation map. According to still further features m the described prefeπed embodiments providing the color karyotype of the malignant cells is for obtaining a color translocation map.

The present invention successfully addresses the shortcomings of the presently known configurations by providing a metiiod for in situ hybridization which is sensitive enough to simultaneously detect dozens of spectrally similar, yet some what different fluorescent probes, thus, the metiiod of the present invention is capable of providing a color karyotype in which each chromosome pair appears in a different RGB or artificial color; simultaneous loci mappmg of dozens of loci; a combination of color karyotyping and multiple loci mapping and readily available genetic inspection for chromosomal abeπations.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention herein described, by way of example only, with reference to the accompanying drawings, wherein:

FIG. 1 illustrates a conventional (prior art) slit-type imagmg spectrometer;

FIG. 2 is a block diagram illustrating the main components of an imaging spectrometer constructed in accordance with U.S. Pat application No. 08/392,019 (prior art); FIG. 3 illustrates a non-moving type interferometer, namely, a Sagnac interferometer, as used in an imaging spectrometer in accordance with U.S. Pat application No. 08/392,019 (prior art);

Fig 4 shows a definition of pseudo-RGB (Red, Green and Blue) colors for emphasizing chosen spectral ranges The intensity for each pseudo-color is calculated by integrating the area under the curve, after multiplying it by one of the curves FIGs 5a, 5b and 5c show interphase FISH performed with two different probes attached to Texas-Red and Rhodamine wherem (a) is an ongmal unage, the way it looks thorough a microscope, (b) is the same sample, after being measured and processed by the method of the present invention, and (c) are the fluorescence spectra of the Texas-Red and Rhodamine fluorophores, FIGs 6a, 6b and 6c show interphase FISH performed with six different probes each labeled with a different fluorophore wherem (a) is an ongmal image, the way it looks tiiorough a microscope, cells were counter stamed wrth DAPI, (b) is the same sample, after bemg measured and processed by the method of the present invention, and (c) are the fluorescence spectra of the six fluorophores, FIGs 7a, 7b, 7c, 7d and 7e collectively show 24 normalized spectra of 24 pixels of the image of Figures 8a and 9a, each of the 24 pixels is denved from a different human chromosome ( 1-22, X and Y), each of the chromosomes was pamted usmg a different chromosome pamt as detaded m Tables 3 and 4 below ,

FIGs 8a and 8b are an RGB image and a color karyotype (presented m black and white) denved from it, respectively, of the 24 human male chromosomes (1-22, X and Y) each of the chromosomes was pamted usmg a different chromosome pamt as detailed in Tables 3 and 4 below, obtained usmg die method of the present invention,

FIGs 9a and 9b are color presentations of Figures 8a and 8b, respectively, FIGs 10a and 10b are a DAPI R-banding photograph obtamed with a conventional fluorescence microscope and an RGB color karyotype obtamed usmg the chromosome pamts as m Figures 9a-b and the metiiod of the present invention (presented m black and white), respectively, of a female breast cancer cell chromosome spread, FIGs 11a and l ib are an ongmal and more clear presentation and a color presentation of Figure 10a and 10b, respectively,

FIG 12 is a comparative presentation of the DAPI R-banded and an mterpreted oudme of the pamted translocated chromosomes presented m Figures 1 1a (left) and l ib (nght), respectively, as was determined from Figure l ib and mterpreted usmg the color karyotype shown m Figure 9b, and FIGs 13 is a color presentation of Figure 12

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention is of a spectral imagmg method for detecting and analyzing fluorescent in situ hybπdizations employmg numerous chromosome pamts and or loci specific probes each labeled witii a different fluorophore or a combmation of fluorophores, the method is highly sensitive both m spatial and spectral resolutions and is capable of simultaneous detection of dozens of flourophores or combinations of flourophores, therefore, the method of me present invention can be used for the detection of fluorescently pamted complete sets of chromosomes and/or multiple loci from a species such as human and to provide a color karyotype

For purposes of better understanding die present invention, as illustrated m Figures 4-9 of the drawings, reference is first made to the construction and operation of a conventional (l e , pnor art) slit-type imagmg spectrometer utilizing a two-dimensional aπay of detectors as illustrated m Figure 1 Thus, the pnor art slit-type imagmg spectrometer as dlustrated m Figure 1 compπses a collection optical system as indicated at 2, for collecting the mcident light from a scene, schematically mdicated at 4 and focusmg die substantially parallel light of the scene 4 onto a first focal plane occupied by a s t 6 to define the field of view The hght exiting from slit 6 is collimated m a collimator lens 8 and is passed d rough a spectral dispeision element 10 (e g , a grating) to separate the vaπous wavelengths The output from spectral dispersion element 10 is focused by a focusmg lens 12 onto a two-dimensional detector aπay 14 m a second focal plane The output of detector array 14 is fed to a signal processor 16

In the two-dimensional aπay of detectors 14 illustrated m die pnor art imagmg spectrometer of Figure 1, the movement of the system (e g , a raster movement or lme scanning mdicated by arrow 13) effects the scanning along one dimension The scanning along the second dimension is effected by the slit 6 which is oπented perpendicularly to the direction of movement of the system The slit 6 thus assures that each detector withm the aπav 14 sees onlv the contnbution of one pixel at a smgle wavelength at any time This is necessary to separate the spectra of each pixel

As mentioned m d e background section and heremabove, the disadvantage of d e pnor art method illustrated m Figure 1 is that most of the pixels of one frame are not measured at any given tune even though the optical system 2 actually collects energy from all of them simultaneously As a result, the requued frame time is significantly mcreased, and/or the signal-to-noise ratio (sensitivity) is substantially decreased with respect to a system which does not have me need for such a sht

Figure 2 is a block diagram dlustrating the main components of an improved prior art imaging spectrometer disclosed in U.S. Pat application No.

08/392,019 to Cabib et al., filed Feb. 21st, 1995 which is incorporated by reference as if fully set forth herein. This imaging spectrometer is constructed highly suitable to implement the methods of the present invention.

Thus, d e prior art imaging spectrometer of Figure 2 includes: a collection optical system, generally designated 20; a one-dimensional scanner, as indicated by block 22; an optical patii difference (OPD) generator or mterferometer, as indicated by block 24; a one-dimensional or two-dimensional detector aπay, as indicated by block 26; and a signal processor and display, as indicated by block 28.

A critical element in system 20 is the OPD generator or interferometer 24, which outputs modulated hght coπesponding to a predetermined set of linear combinations of the spectral intensity of the light emitted from each pixel of the scene to be analyzed. The output of d e interferometer is focused onto die detector aπay 26. Thus, all the required optical phase differences are scanned simultaneously for all the pixels of the field of view, in order to obtain all the information required to reconstruct the spectrum. The spectra of all the pixels in die scene are thus collected simultaneously witi the imaging information, thereby permitting analysis of the image in a real-time manner.

The apparatus according to U.S. Pat. application No. 08/392,019 may be practiced in a large variety of configurations. Specifically, the interferometer used may be combined with oti er minors as described in the relevant Figures of U.S. Pat. application No. 08/392,019. Thus, accordmg to U.S. Pat. application No. 08/392,019. alternative types of interferometers may be employed. These include (1) a moving type interferometer in which the OPD is varied to modulate the hght, namely, a Fabry-Perot interferometer with scanned tiiickness; (2) a Michelson type interferometer which includes a beamsplitter receiving the beam from an optical collection system and a scanner, and splitting the beam into two paths; (3) a Sagnac interferometer optionally combined with other optical means in which interferometer the OPD varies with the angle of incidence of die incoming radiation, and (4) a four-mirror plus beamsplitter interferometer as further described in the cited U.S. Pat. application. Figure 3 illustrates an imaging spectrometer constructed in accordance with

U.S. Pat. application No. 08/392,019 utilizing an interferometer in which the OPD varies with d e angle of incidence of the mcoming radiation. A beam entering the

interferometer at a small angle to the optical axis undergoes an OPD which varies substantially linearly with this angle.

In die interferometer of Figure 3, all the radiation from source 30 in all the pixels, after being collimated by an optical collection system 31, is scanned by a mechanical scanner 32. The light is then passed tiirough a beamsplitter 33 to a first reflector 34 and tiien to a second reflector 35, which reflects the hght back through the beamsplitter 33 and tiien through a focusing lens 36 to an aπay of detectors 37 (e.g., a CCD). This beam interferes with the beam which is reflected by 33, then by second reflector 35, and finally by first reflector 34. At the end of one scan, every pixel has been measured tiirough all the

OPD's, and therefore the spectrum of each pixel of the scene can be reconstructed by Fourier transformation. A beam parallel to the optical axis is compensated, and a beam at an angle (θ) to the optical axis undergoes an OPD which is a function of the diickness of d e beamsplitter 33, its mdex of refraction, and the angle θ. The OPD is proportional to ( for small angles. By applying the appropriate inversion, and by careful bookkeeping, the spectrum of every pixel is calculated.

In the configuration of Figure 3 the ray which is incident on d e beamsplitter at an angle β (β = 45° in Figure 3) goes through the interferometer with an OPD = 0, whereas a ray which is incident at a general angle β - θ undergoes an OPD given by the following:

0?D(β, θAn) (A-sm ² (β-θ) ⁵ + 2sιn/?sιn£] ( 1)

where β is the angle of mcidence of the ray on the beamsplitter; θ is the angular distance of a ray from the optical axis or mterferometer rotation angle with respect to the central position; t is the diickness of the beamsplitter; and n is the index of refraction of the beamsplitter.

It follows from Equation 1 that by scanning both positive and negative angles witii respect to the central position, one can get a double-sided interferogram for every pixel, which helps eliminate phase eπors giving more accurate results in the Fourier transform calculation. The scanning amplitude determines the maximum OPD reached, which is related to the spectral resolution of the measurement. The size of the angular steps deteπnines the OPD step which is, in turn, dictated by the shortest wavelength to which the system is sensitive. In fact, according to die sampling theorem [see, Chamberlain (1979) The principles of interferometric spectroscopy, John Wdey and Sons, pp. 53-55], this OPD step must be smaller than half the shortest wavelength to which the system is sensitive.

Another parameter which should be taken into account is the finite size of a detector element m the matrix Through the focusmg optics, the element subtends a finite OPD in the interferometer which has the effect of convolving die mterferogram with a rectangular function This brings about, as a consequence, a reduction of system sensitivity at short wavelengths, which drops to zero for wavelengths equal to or below the OPD subtended by the element For this reason, one must ensure that the modulation transfer function (MTF) condition is satisfied, 1 e , tiiat die OPD subtended by a detector element m the mterferometer must be smaller than the shortest wavelength at which the instrument is sensitive Thus, imaging spectrometers constructed m accordance with d e invention disclosed m U.S Pat. application No 08/392,019 do not merely measure the intensity of light coming from every pixel m the field of view, but also measure the spectrum of each pixel in a predefined wavelength range. They also better utilize all the radiation emitted by each pixel m the field of view at any given time, and therefore permit, as explamed above, a significant decrease in d e frame time and or a significant increase in the sensitivity of the spectrometer Such imagmg spectrometers may include vaπous types of interferometers and optical collection and focusmg systems, and may therefore be used m a wide vaπety of applications, mcludmg medical diagnostic and therapy and biological research applications, as well as remote sensmg for geological and agπcultural investigations, and d e like. An imagmg spectrometer m accordance witii die invention disclosed m U.S Pat. application No 08/392,019 was developed by Spectral Diagnostics (SD) Ltd , Industrial Park, Migdal Haemek. Israel and will be refeπed herembelow as SpectraCube™ The SpectraCube™ system optically connected to a vaπety of optical devices was used to implement the method of d e present invention The

SpectraCube™ system has the following characteristics, hsted herembelow m

Table 1.

TABLE 1 :

Character Performance

Spatial resolution: 30/M μm (M=effective microscope or fore optics magnification)

Field of View 8/M millimeter

Sensitivity: 20 milliLux (for 100 msec integration time, mcreases for longer integration times linearly with )

Spectral range ^• 400- 1000 nm

Spectral resolution: 4 nm at 400 nm (16 nm at 800 nm)

Acquisition time: 5-50 sec, typical 25 sec

FFT processing time: 20-180 sec, typical 60 sec

DISPLAY AND ANALYSIS OF SPECTRAL IMAGES

a. General

As mentioned above, a spectral image is a three dimensional aπay of data,

I(x,y,λ), that combines spectral information with spatial organization of d e image. As such, a spectral image is a set of data called a spectral cube, due to its dimensionality, which enables die extraction of features and d e evaluation of quantities that are difficult, and in some cases even impossible, to obtain otherwise. Since both spectroscopy and digital image analysis are well known fields that are covered by an enormous amount of literature [see, for example, Jain ( 1989) Fundamentals of Digital Image Processing, Prentice-Hall International], the following discussion will focus primarily on die benefit of combining spectroscopic and imaging information in a single data set i.e., a spectral cube. One possible type of analysis of a spectral cube is to use spectral and spatial data separately, i.e., to apply spectral algorithms to the spectral data and two- dimensional image processmg algorithms to the spatial data.

As an example of a spectral algorithm, consider an algorithm computing die similarity between a reference spectrum and d e spectra of all pixels (i.e., similarity mapping) resulting in a gray (or other color) scale image (i.e., a similarity map) in which the intensity at each pixel is proportional to the degree of 'similarity'. This gray scale image can then be further analyzed usmg image processing and computer vision techniques (e.g.. image enhancement, pattern recognition, etc.) to extract the desired features and parameters. In other words, similarity mapping involves computing the mtegral of the absolute value of the difference between the spectrum of each pixel of the spectral image with respect to a reference spectrum (either previously memorized in a library, or belonging to a pixel of the same or other spectral image), and displaying a gray level or pseudocolor (black and white or color) image, in which the bright pixels coπespond to a small spectral difference, and dark pixels coπespond to a large spectral difference, or vice versa.

Similarly, classification mapping perform the same calculation as described for similarity mapping, yet takes several spectra as reference spectra, and paints each pixel of the displayed image with a different predetermined pseudocolor, according to its classification as being most similar to one of the several reference spectra.

It is also possible to apply spectral image algorithms based on non- separable operations; i.e., algorithms that include botii local spectral information

and spatial coπelation between adjacent pixels (one of tiiese algorithms is, as will be seen below, a principal component analysis).

One of the basic needs that arise naturally when dealing witii any tiiree- dimensional (3D) data structure such as a spectral cube (i.e., I(x,y,λ)), is visualizing tiiat data structure in a meaningful way. Unlike other types of 3D data such as tomographic data, D(x,y,z), obtained for example by a confocal microscope, where each point represents, in general, the intensity at a different locations (x,y,z) in tree-dimensional space, a spectral image is a sequence of images representing the intensity of the same two-dimensional plane (i.e., die sample) at different wavelengths. For this reason, the two most intuitive ways to view a spectral cube of data is to either view the image plane (spatial data) or the intensity of one pixel or a set of pixels as function of wavelength in a three- dimensional mountain-valley display. In general, the image plane can be used for displaying eiti er die intensity measured at any single wavelength or the gray scale image that results after applying a spectral analysis algorithm, over a desired spectral region, at every image pixel. The spectral axis can, in general, be used to present the resultant spectrum of some spatial operation performed in die vicinity of any desired pixel (e.g., averaging the spectrum).

It is possible, for example, to display d e spectral image as a gray scale image, similar to the image that might be obtained from a simple monochrome camera, or as a multicolor image utilizing one or several artificial colors to highlight and map important features. Since such a camera simply integrates the optical signal over the spectral range (e.g., 400 nm to 760 nm) of the CCD aπay, the 'equivalent' monochrome CCD camera image can be computed from d e 3D spectral image data base by integrating along the spectral axis, as follows:

12 gray _scale(x,y) = ) w( λ) -I(x,y,λ)dλ. (2) λl

In Equation 2, w(λ) is a general weighting response function tiiat provides maximum flexibility in computing a variety of gray scale images, all based on d e integration of an appropriately weighted spectral image over some spectral range. For example, by evaluating Equation (2) with three different weighting functions, {w λj, w (λ), w _h (λ) , coπesponding to die tristimulus response functions for red (R), green (G) and blue (B), respectively, it is possible to display a conventional RGB color image. It is also possible to display meaningful non-conventional (pseudo) color images. Figure 4 presents an example of the power of this simple algorithm. Consider choosing {w w„ \v _b ] to be Gaussian functions distributed "inside" a spectrum of interest, the resulting pseudo-color image that is displayed

in this case emphasizes only data in the spectral regions coπesponding to die weighting functions, enabling spectral differences in these three regions to be detected more clearly.

b. Point operations

Point operations are defined as tiiose that are performed on single pixels,

(i.e., do not involve more than one pixel at a time). For example, in a gray scale image, a point operation can be one that maps me intensity of each pixel (mtensity function) into another intensity accordmg to a predetermined transformation function. A particular case of this type of transformation is the multiplication of the intensity of each pixel by a constant.

The concept of point operations can also be extended to spectral images: here each pixel has its own intensity function (spectrum), i.e., an n-dimensional vector V ( ): Λe [ .„ A point operation applied to a spectral image can be defined as one tiiat maps me spectrum of each pixel into a scalar (i.e., an intensity value) according to a transformation function:

v ₂ = g(V,(λ))A^ [λ _x , λ, (3)

Budding a gray scale image according to Equation 3 is an example of this type of point operation. In the more general case, a point operation maps the spectrum (vector) of each pixel into another vector according to a transformation function:

^( = grø)); /e [ l, N], λe [λ„ ΛJ (4), where N < n.

In this case a spectral image is transformed into another spectral image.

One can now extend d e definition of point operations to include operations between coπesponding pixels of different spectral images. An important example of this type of algorithm is optical density analysis. Optical density is employed to highlight and graphically represent regions of an object being studied spectroscopically with higher dynamic range than the transmission spectrum. The optical density is related to transmission by a logarithmic operation and is therefore always a positive function. The relation between the optical density and die measured spectra is given by Lambert Beer law:

Ou(A -\ _ω -^- --\^ rl^ (5)

where OD(λ) is the optical density as a function of wavelength, I(λ) is the measured spectrum, I ₀ ( l) is a measured reference spectrum, and τ(λ) is the spectral transmitance of the sample Equation 5 is calculated for every pixel for every wavelength where I ₀ {λ) is selected from (1) a pixel m the same spectral cube for which OD is calculated, (2) a coπesponding pixel m a second cube, and (3) a spectrum from a library

Note that the optical density does not depend on e ther die spectral response of the measuring system or the non-uniformity of the CCD detector This algonthm is useful to map the relative concentration, and m some cases the absolute concentration of absorbers m a sample, when their absorption coefficients and d e sample diickness are known It should tiius be noted tiiat the term 'level' as used herem also refers to the terms 'amount', 'relative amount', 'absolute concentration' and 'relative concentration' Additional examples mclude vanous lmear combmation analyses, such as for example ( 1) applymg a given spectrum to the spectrum of each of the pixels m a spectral image by an aπthmetical function such as addition, subtraction, multiplication division and combinations thereof to yield a new spectral cube, m which the resulting spectrum of each pixel is the sum, difference, product ratio or combmation between each spectrum of the first cube and d e selected spectrum, and (2) applymg a given scalar to the spectra of each of the pixels of the spectral unage by an anthmetical function as descπbed above

Such lmear combinations may be used, for example, for background subtraction m which a spectrum of a pixel located in d e background region is subtracted from die spectrum of each of the pixels, and for a cahbration procedure m which a spectrum measured pnor to sample analysis is used to divide d e spectrum of each of the pixels m the spectral image

Another example mcludes a ratio unage computation and display as a gray level image This algorithm computes the ratio between the mtensities at two different wavelengths for every pixel of the spectral image and pamts each of the pixels m a lighter or darker artificial color accordmgly For example, it pamts d e pixel bπght for high ratio, and dark for low ratio (or the opposite), to display distributions of spectrally sensitive matenals

c. Spatial-spectral combined operations

In all of the spectral image analysis methods mentioned above, algontirms are applied to die spectral data The importance of displaying die spectrally

processed data as an image is mostly qualitative, providing die user with a useful image. It is also possible, however, depending on die application, to use the available imaging data in even more meaningful ways by applying algorithms that utilize the spatial-spectral coπelation that is inherent in a spectral image. Spatial- spectral operations represent the most powerful types of spectral image analysis algorithms. As an example, consider die following situation:

A sample contains k cell types stained witii k different fluorophores (the term 'cell' here is used botii for a biological cell, and also as 'a region m the field of view of the instrument'). Each fluorophore has a distinct fluorescence emission spectrum and binds to only one of the k cell types. It is important to find die average fluorescence intensity per cell for each one of the k cell types. To achieve this task the following procedure can be used: (1) classify each pixel in the image as belonging to one of k+l classes (k cell types plus a background) according to its spectrum; (2) segment the image into the various cell types and count the number of cells from each type; and (3) sum the fluorescence energy contributed by each class, and divide it by die total number of cells from the coπespondmg class.

This procedure makes use of both spectral and spatial data. The relevant spectral data takes d e form of characteristic cell spectra (i.e., spectral "signatures"), while the spatial data consists of data about various types of cells (i.e., cell blobs) many of which appear similar to the eye. The ideal type of measurement for this type of situation is a spectral image. In the above situation, cells can be differentiated by their characteristic spectral signature. Hence, a suitable point operation will be performed to generate a synthetic image in which each pixel is assigned one of k+ l values. Assuming that die fluorescence emission spectra of the different cell types are known to be s _{ (λ) i = 1. 2, , k, λ λ , v , and die measured spectrum at each pixel (x, y) is s _x (λ), λe [λ _x , ΛJ, then die following algorithm is a possible method of classification (step 1 above):

Let e ¹ _l be the deviation of the measured spectrum from the known spectrum of the fluorophore attached to cell type / ^' . Then, adopting a least-squares "distance" definition, one can write:

where R _λ is the spectral region of interest. Each point [pixel (x, y)] in the image can then be classified into one of the k- \ classes using die following criterion:

-po t(x,y) e class k+l if e-j > threshold for all / e [ l,k], whereas (7) pomt(x,y) e class p if e^ _; - < threshold, and p is such that m [e-i]= e-p

Steps 2 and 3 above (image segmentation and calculation of average fluorescence intensity) are now straight-forward using standard computer vision operations on die synthetic image created in accordance with the algorithm described in Equations 6 and 7. Another approach is to express the measured spectrum s _{x v} (λ) at each pixel as a linear combination of the k known fluorescence spectra s,(λ); i = 1, 2, , k.

In this case one would find the coefficient vector C = [c _] , c ₂ , , c _k \ that solves:

= min ∑(s(λ) -s(λ)) ² (8) λeR k where s(λ) = ∑c _t -s _t (λ), i-l dF

Solving for — =0;for ι - l,2 k (ι.e., find values of c which πunimize F) dc, where A is a square matrix of dimension

( 10), and B is a vector defined as i_ = ^■ M m,n = ,2, .. „k ( 1 1). Arithmetic operations may similarly be applied to two or more spectral cubes and or spectra of given pixels or from a library. For example consider applying an arithmetic operations between coπesponding wavelengths of coπesponding pairs of pixels belonging to a first spectral cube of data and a second spectral cube of data to obtain a resulting third spectral cube of data for the purpose of, for example, averaging two spectral cubes of data, time changes follow-up, spectral normalization, etc.

In many cases objects (e.g., cells) present in a spectral image differ from one another in chemical constituents and/or structure to some degree. Using a principal component analysis by producing covariance or coπelation matrix enhances these small differences. A brief description of the prmcipal component analysis using a covariance matrice is given below. For further detads regarding die principal component analysis, the reader is refeπed to Martens and Naes (1989) Multivariate Cahbration, John Wiley & Sons, Great Britain; and to Esbensen et al., Eds. (1994) Multi variance analysis - in practice. Computer-aided modeling as CAMO, and die Unscrambler's User's guide, Trondheim, Norway.

Thus, the intensities of the pixels of the image at wavelength λ[ (i = 1,...N) are now considered a vector whose length is equal to the number of pixels q. Since there are N of these vectors, one for every wavelength of d e measurement, these vectors can be aπanged in a matrix B' with q rows, and N columns:

No. of wavelengths B ^{• ■} • _v ^

B' = No. of pixels • (12)

B ^ B qH

For each of the columns of matrix B' defined is an average:

. ₌ - ∑ B\. ; / = / .N (13)

and a second normalized matrix B defined as'

No. of wavelengths

B u. B \ ^" tt. , ' ^" N

B No. of pixels (14)

A covariance matrix C is defined for the matrix B: C = B^-B of dimensions

NxN. C is diagonahzed, and eigenvectors and eigenvalues related by: C = μ\- V, where Vi axe Ν orthogonal unit vectors and μj axe the eigenvalues representing the variance in the direction of the ;-th unit vector V,. In general, the lowest components represent the highest variability as a function of pixels. The products BVj (i = 1,...N) are the projections of the spectral image onto the elements of the orthogonal basis, They are vectors with q elements (q = number of pixels), and can be displayed separately as black and white images. These images may reveal features not obvious from a regular black and white image filtered at a certain wavelength or wavelength range.

FLUORESCENCE MICROSCOPY:

a. General

The use of multiple dyes (i.e., fluorophores) [see, Jain ( 1989) Fundamentals of Digital Image Processing, Prentice-Hall International], is one of the most powerful and common tools for analyzing tissues and cells. Fluorescence microscopy is therefore one of the most important experimental methods used in light microscopy [Lakowicz ( 1983) Principles of fluorescence spectroscopy, Plenum Press, New York, London]. The power of fluorescent probes is mainly due to d e great variety of biological structures to which specific dyes can be bound [Waggoner ( 1986) Applications of fluorescence in the biomedical sciences, Eds. Taylor et al., New York: Alan R. Liss, Inc. pp. 3-28]. For a detailed review of fluorescent probes see, Mason (editor) ( 1993) Fluorescent and Luminescent Probes for Biological Activity, Biological Techniques Series, edited by Sattelle, Academic Press Limited, London; and, Ploem and Tanke (1987) Introduction to Fluorescence Microscopy, Oxford University Press, Royal Microscopical Society.

The rapid development of new and more sophisticated multicolor fluorescent dye molecules wdl continue to create a need for more advanced fluorescence imaging techniques that can utilize the full potential of these dyes. For a discussion of the revolutionary impact fluorescent dyes have had, and will continue to have, on the way research is conducted today, refer to Taylor et al. (1992) The New Vision of Light Microscopy, .American Scientist, Vol. 80, pp. 322-335. A remarked improvement in multicolor fluorescent dyes is die introduction of combinatorial fluorescent dyes which are various combinations of few basic fluorescent dyes, see, Ried et al, (1992) Simultaneous visualization of seven different DNA probes by in situ hybridization using combinatorial fluorescence and digital imaging microscopy. Proc. Natl. Acad. Sci. USA 89, 1388-1392; and, Ried (Jan. 1994) Fluoreszenz in situ Hybridizierung in der genetischen Diagnostik, Faculty of theoretical medicine, Ruprecht-Karls University Heidelberg.

Spectral bio-imaging using the method of the present invention, provides several important advantages for fluorescence imaging applications over simple filter based approaches. These advantages include d e following: (1) measurement of die complete spectrum, providing much more quantitative insight into the actual behavior of dye molecules in the sample of interest; (2) ability to overcome many of the traditional problems arising from undesirable background luminescence; (3)

undesirable or unpredictable spectral shifts that occur in die emission spectrum of a fluorescent probe, due to its micro-environment (e.g., temperature), can be taken into account in determining the probe concentration, whereas when the fluorescence intensity is only measured with a band-pass filter, such spectral shifts would not only go undetected but might cause significant eπors in analyzing the probe concentration; and, (4) simplification of fluorescence image acquisition and, as will be shown below in detail, when used in conjunction with the appropriate spectral analysis algorithms it is possible to separate and map, in a single measurement, many spectrally overlapping fluorescent dyes. In fact, by applying sophisticated data analysis algorithms such as multivanate analysis, principal component regression and otiier classification algorithms [see, Martens and Naes ( 1989) Multivariate Cahbration, John Wiley & Sons, Great Britain] it is possible to analyze many spectrally related parameters simultaneously.

Spectral bio-imaging accordmg to the present invention provides means for eliminating problems associated with undesirable background luminescence as follows. Fluorescence imaging microscopy is typically performed by usmg a fluorescence filter cube which ensures that d e sample is excited by die desired short wavelengths, and tiiat only wavelengths in a limited spectral band coπesponding to the fluorescence emission of the probe reach the detector (e.g., eye, camera, etc.) [Mason (editor) (1993) Fluorescent and Luminescent Probes for Biological Activity, Biological Techniques Series, edited by SatteUe, Academic Press Limited, London]. Since fluorescence intensities are usually several orders of magnitude below d e intensity of the excitation source, such background luminescence can never be eliminated perfectly [Benson et al. ( 1985) Cell Biol. 100, pp. 1309-1323]. The three primary sources for undesirable background luminescence are: (1) radiation from the excitation source that is not completely blocked by d e dichroic minor coating and or die filter; (2) auto-fluorescence of the sample, and sometimes also from the optical elements; and (3) selection of an inappropriate (or sub-optimal) combination of excitation filter, dichroic minor and barrier filters. These sources can contribute significantly to the background fluorescence. The effects of sample auto-fluorescence can usually be reduced by selecting fluorescent probes whose absoφtion and emission bands do not overlap with tiiose of the sample being measured. Similarly, by choosing optical elements tiiat are appropriately coated to reduce auto-fluorescence, the effects of this type of auto-fluorescence can also be minimized.

In spite of the best filtering methods available, undesirable background luminescence makes it often difficult, and sometimes impossible, to bring out the relevant fluorescence signal from its background (noise). The spectral bio-imaging

method of d e present mvention is able, on the otiier hand, to use spectral differences between (1) the spectral shape and spectral range of the fluorescent dye and (n) d e spectral shape and spectral range of the background luminescence (mcludmg auto-fluorescence), to eliminate the effects of undesuable background luminescence

Thus, by applymg the appropnate spectral image analysis methods to the emission spectra of fluorescent probes, it is possible to improve the signal-to-noise ratio, and hence die accuracy, of fluorescence imaging measurements This advantage of the spectral bio-imaging approach is of particular importance for ratio imagmg, when quantitation of the results is desued In addition, the spectral bio¬ imaging system of the present mvention can save time and effort that is otiierwise spent m choosmg the optimal filters for a filter based measurement

The acquisition of multicolor fluorescence images can be greatly simplified when die power of spectral bio-imaging accordmg to the method of the present mvention, is combmed witii the appropnate fluorescent markers In order to fully realize the benefits afforded by spectral bio-imagmg, the reader is asked to consider d e typical steps mvolved in usmg a filter based imagmg method to measure the fluorescence from a sample containing multiple probes Fust, probes with sufficiently different absorption and emission spectra must be selected In todays practice, this requirement limits the number of fluorescent markers m a specimen to between three and five probes Fluorescence images are then acquired, one unage for each dye, by appropnately rotating two filter wheels, one for selecting the excitation wavelength and anotiier for capturing die emission spectrum, or alternatively rotating one filter wheel aimed at selecting die excitation wavelength, while capturing die emission spectrum is by a triple dichroic filter Approaches in which tunable filters (no moving parts) are used to control the excitation and or emission wavelength have also been proposed Recently, multispectral interference filters have also been used to enable imaging multiple fluorophores [Lengauer et al ( 1993) Human Molecular Genetics 2, pp 505-512] Means of changmg the dichroic minor (e g , by changmg filter cubes) is also requued It is also frequently necessary to readjust d e focus of the unage at each wavelength and sometimes even the CCD camera exposure time must be changed to achieve higher signal-to-noise ratios Collectively, these limitations create a registration problem The resulting monochrome images, each coπespondmg to die emission of a different fluorescent dye, are then pseudo- colored and superimposed (usmg a digital computer with readily avadable off-the- shelf software) The resulting unage shows the location of several fluorescent markers, each colored with a different pseudo-color Smce slight changes m the

position of the dichroic minor will cause translational shifts in the digitized images, it is necessary to use multiple wavelength dichroic minors [for use of a dichroic with quadruple wavelength band-pass properties see. Hiraoka et al. (1992) Seminars in Cell Biology, Vol. 2, pp. 153-164] or to register the images prior to their superposition. The image registration approach is more common, despite die fact that image registration is a difficult problem which can be time consuming and often produces only marginally satisfactory results. These are technical challenges which must also be addressed when acquiring multi-color fluorescence images [ Waggoner et al. ( 1989) Part B of Methods in Cell Biology, Vol. 30, Ch. 17, pp. 449-478, edited by Taylor and Wang, Academic Press Inc.].

The spectral bio-imaging method of the present invention thus overcome one of the fundamental limitations imposed by filter based approaches to fluorescence imaging. By enabling die simultaneous measurement of the emission spectrum of an unlimited number of fluorescent dyes (including dyes whose emission spectra overlap to a great extent, as demonstrated hereinbelow in the Examples section for the Texas-Red and Rhodamine fluorophores), spectral bio¬ imaging eliminates the need for sequentially acquiring images of the emissions of multiple fluorescent probes. The advantage of using a spectral bio-imaging system is greatest when the used fluorescent probes can be excited by a common excitation source. In this case, a single spectral image acquisition can capture the fluorescence emission of an almost unlimited number of dyes and die need to (1) select non-overlapping dyes; (2) change filter cubes; (3) change excitation or emission filters; (4) optimize the focus and/or exposure time or (5) register the images, is eliminated. The challenge, of course, is to select suitable dyes that can be excited with a common source. Dyes which are excited by fluorescence energy that is transfeπed to/from one another are thus ideally suited for multi-color fluorescence imaging using a spectral bio-imaging system. Clearly, the use of dyes witii similar emission properties will make visual detection (e.g., under the microscope) more difficult; however, this limitation is likely to be solved using the spectral bio-imaging method of the present invention.

b. Spectral identification of multiple fluorophores

The use of the spectral bio-imaging method according to the present invention enables the simultaneous measurement of many dyes (i.e., fluorophores, fluorescent moieties) in one measurement. There is no restriction on the type of dye, even dyes tiiat overlap spectrally (e.g., Rhodamine and Texas-Red) can be identified as will be exemplified below (see, Example 1 and 2) by applying

suitable algonthms (e g , lmear combmation for background subtraction, etc ) and their occuπence mapped m an unage However, if many dyes are to be used simultaneously , careftd consideration should be given to their excitation wavelengths, fluorescence mtensities and emission spectra When this is done properly, the results can be analyzed quantitatively as well For example, the relative concentration of several proteins can be mapped m a smgle measurement usmg suitable fluorescentiy tagged antibodies which specifically bmd to these protems By usmg standard cahbrated dyes, die absolute concentrations can also be determined One important example where the detection of multiple fluorescent probes can be a significant advantage is FISH (fluorescent in situ hybπdization) [Emanuel (1993) Growth Genetics and Hormones 9, pp 6-12], which is used to analyze genes at the chromosome level, and find possible genetic defects such as gene/chromosome amplification, deletion, translocation, reaπangement and other abnormalities

Certain diseases and disorders, mcludmg many cancers and birth defects, are genetic disorders caused by defects m one or more genes Many other diseases are known or believed to have a genetic component's), mat is, there exists genetic defect(s) tiiat does not alone cause the disease but contnbutes to it, or increases the probability of developmg die disease later m life, phenomena known m the art as multifactonal diseases and genetic predispositions Coπelation of visible genetic defects witii known diseases would allow doctors to make definitive diagnoses, and permit early detection and treatment of many diseases Genetic counseling could alert prospective parents and at-πsk individuals to the possibdity of potentiahy senous medical problems m the future, permitting appropnate intervention

More than 5,000 genetic disorders have now been identified, many of which are associated witii multiple genetic defects After the discovery that chromosomes are the earners of hereditary information, scientists reasoned that it should be possible to document visible defects m chromosomes that were responsible for specific disorders In the 1960's, staining techniques were developed for microscopy-based classification of metaphase chromosomes spread onto glass slides For several decades, visual analysis of chromosomes banding patterns has been used to coπelate human genetic disorders with observed structural abnormalities m metaphase chromosomes Chromosomes are typically examined by bπghtfield microscopy after Giemsa staining (G-banding), or examined by fluorescence microscopy after fluorescence staining (R-banding), to reveal characteπstic light and dark bands along tiierr length Careful companson

of a patient's banding pattern with those of normal chromosomes can reveal abnormalities such as translocations (exchange of genetic material between or within chromosomes), deletions (missing chromosomes or fragments of chromosomes), additions, inversions and otiier defects that cause deformities and genetic diseases.

However, many serious genetic diseases, such as for example cystic fibrosis (CF) and many others, are caused by mutations that involve addition, deletion or substitution of only one or a few nucleotides. Such small defects are not detectable by die chromosomal banding techniques described above, and for many years cytogeneticists have been working to develop techniques for locating and quantifying minute defects.

Fluorescent in situ hybridization (FISH) has evolved over the past 25 years through the improvement of a number of complementary techniques. Its emergence has been driven by the desire of cytogeneticists to develop better tools for mapping the precise location of genes on chromosomes, and to detect very small genetic defects not visible by gross staining of chromosomes. The human genome project (HGP), a bold initiative to identify and map all human genes, has identified interest in FISH and has hastened die development of much-needed DNA probes. Cunent FISH techniques have also been made possible by die concunent development of powerful immunological probes, a growing variety of excellent fluorescent dyes for microscopy and spectroscopy, and dramatic improvements in the objectives, illuminators and filters used for fluorescence microscopy.

The power and utility of FISH is due to many factors: ( 1) FISH can be used not only on isolated chromosomes and nuclei, but also whole cells witiiin fixed, paraffin-embedded tissue sections; (2) it can detect relatively small defects (ability of detecting smaUer defects being constantly increased); (3) it can provide results relatively quickly; (4) its moderate cost allows it to be used in most diagnostic and research laboratories; (5) adaptation can be developed for various probes and specimen types; and, (6) high specificity and sensitivity can be achieved (7) witiiin a short throughput, typically two hours.

Many FISH applications require only that die cytogeneticist look tiirough the eyepieces of a microscope, or at the image on the monitor, to determine whether a fluorescent label is present or absent. With somewhat more complex specimens, a simple count of one or two colored labels may be done. However, the ability to process digital images and extract numerical data from them adds a vast new set of capabdities to FISH techniques. An appropriate imaging method, such as die method of the present invention, can enhance very faint FISH images

so that labeled chromosomes and loci are clearly identifiable Under readdy achieved experimental conditions, die number of labeled sites can be automatically counted In addition, die mtensity at each labeled site can be measured and die amount of DNA calculated to reveal, for example, the number of copies present of a particular gene Emergmg techniques such as multicolor FISH employ color image analysis to detect and quantify multiple (3,4,5 and more) fluorescent probes

As discussed above, FISH can provide information on the location of the labeled probe, the number of labeled sites on each chromosome, and the mtensity of labeling (the amount of genetic mateπal) at each site Centromeπc (repetitive DNA) probes and chromosome pamts are used to tag and count the number of copies present of each targeted chromosomes Locus-specific probes are used to map e location of smad regions of genetic matenal These types of probes can be used on mtact interphase nuclei as well as metaphase chromosome spreads, and can be counted visually or automatically by a suitable algonthm They are routinely used to identify genetic diseases characteπzed by having too many or too few copies of a specific chromosome, chromosome fragment, or gene

In very early stages of some cancers, long before the cells are recognizably abnormal, there may be an mcrease m the number of specific genes, phenomenon known m the art as gene amplification, that are detectable usmg locus-specific probes Usmg FISH to detect chromosome abnormalities m cancerous cells may pomt out the developmental stage the disease have reached and therefore to select the most suitable treatment(s), many of which are stage specific m their effectiveness Thereby precious time is saved and patient's suffering is minimized, selecting the most effective stage specific treatment

It is possible to uniformly label the entire surface of one specific chromosome by isolating the chromosome (usmg flow cytometry, for example), physically (e g , by somcation) or errzymatically (e g , by endonucleases) choppmg it up, and generating a set of probes against all of the fragments Whole chromosome probes, also known as chromosome pamts, will fluorescently label all copies of the target chromosome One important application of chromosome painting is the detection of deletions and translocations between two chromosomes, as characteπstically occurs m early stages of certain cancers

For example, if chromosome A is specifically labeled with a green paint and chromosome B is labeled witii a red pamt, any translocation of mateπal from A to B wdl appear as a green area on a red chromosome (and vice versa) Typically, chromosome pamts generated from normal chromosomes are used to detect deletions or translocations on abnormal (patient) chromosomes Reverse

chromosome painting uses probes generated from an abnormal chromosome to identify DNA from various normal chromosomes which contributed material to the abnormal chromosome. The method of die present mvention, as exemplified hereinbelow in the Examples section, enables to pamt the 24 different chromosomes comprising the human karyotype (i.e., genome) each in a different color and simultaneously detect, identify and meaningfully display a color human karyotype, using a single hybridization followed by a single measurement.

Comparative genomic hybridization (CGH) is a variation of reverse chromosome painting in which two cocktails of DNA probes are generated from entire sets of chromosomes. One cocktail is generated from a set of normal chromosomes, and another from a set of abnormal (e.g., tumor) chromosomes. The two sets of probes are generated using different reporter molecules so that, for example, normal DNA will exhibit red fluorescence, and abnormal DNA will exhibit green fluorescence. A normal metaphase spread is hybridized simultaneously with both cocktails, and cunentiy evaluated using color image analysis. Regions of normal chromosomes tiiat fluoresce more mtensely green than red indicate tiiat DNA amplification (multiple gene copies) has occuπed at that gene in the patient's abnormal cells. Regions with more red man green fluorescence (decreased green/red ratio) indicate sites of genetic deletions in the patient's chromosomes, and regions with equal green and red fluorescence indicate tiiat no DNA changes have occuπed at that site. CGH and related techniques are more complex than previous labeling techniques, yet they offer the ability to detect and quantify more subtle and extensive genetic alterations than were previously possible. The method of the present mvention is highly suitable for these types of analyses.

From what has been said above, it follows that karyotypmg, translocation/reaπangement detection, chromosome deletion/amplification, and gene mapping will greatly benefit by the use of the sensitive, quantitative, spectral imaging method of the present invention that builds a whole spectral image at relatively high spectral resolution, instead of a simple color fluorescence image. This is because such method will decrease the sample preparation time and will be able to distinguish between a hybridized fluorescent probe from one that is residual in the background (by small spectral shifts), and will be able to measure a yet not achieved large number of probes, simultaneously. Thus one of the objectives of the present invention is to provide a FISH imaging method designed to exploit die advances in probe technology. According to the present invention there is a possibility of greatly increasing the number of probes that can be analyzed in any given chromosome analysis, as well as

dramatically mcreasmg the speed and degree of automatization at which this information can be acquired as compared witii pnor art methods

The FISH imagmg metiiod of the present mvention exploit the advantages of d e SpectraCube™ system, that is capable of simultaneously acquire fluorescence spectra from all pixels of the microscope field of v ew and detect die location of many fluorescent probes m a smgle experiment In conjunction with the availabdity of chromosome specific probes and novel labeling strategies, and as is exemplified m die Examples below, the metiiod bemg capable of creating a FISH karyotype with each chromosome bemg pamted with a different color (i.e , 24 different colors for a human karyotype) This method result m extremely high sample throughput and allow analysis of essentially unlimited number of probes

As delineated above, the key concepts of the present mvention is the use of many fluorescent probes m FISH assays Numerous methods are avadable to label DNA probes for use m FISH, mcludmg indirect methods whereby a hapten such as biotin or digoxigenin is incorporated mto DNA. usmg enzymatic reactions Following hybndization to a metaphase chromosome spread or interphase nuclei, a fluorescent label is attached to d e hybnd through the use of immunological methods More recently, fluorescent dyes have been d ectly incorporated mto probes and detected without the use of an intermediate step Standard FISH dyes mclude fluorescem, rhodamme, Texas-Red and cascade blue, and multiprobe FISH analysis can be accomplished by labeling different probes with different haptens or fluorescent dyes and combinations thereof, known m the art as combmatonal probes [see, Ried et al , ( 1992) Simultaneous visualization of seven different DNA probes by in situ hybndization usmg combmatonal fluorescence and digital imagmg microscopy Proc Natl Acad Sci USA 89, 1388- 1392, and, Ried (Jan 1994) Fluoreszenz in situ Hybndizierung m der genetischen Diagnostik, Faculty of theoretical medicme, Ruprecht-Karls University Heidelberg]

Fluorescence is a form of luminescence which occurs after photons of light are absorbed by a molecule known as a fluorophore at the ground electronic state The molecule is raised to an excited state as a result of electron transfer to a higher energy orbit This excess energy is dissipated when die electron returns to the ongmal ground state, releasmg a quantum of light The fluorescence light is of longer wavelength than the absorbed light This shift is limited, causmg the emission to be close to the excitation wavelength Because of this, the fluorophores which can be excited m one spectral range emit m a smular spectral range For example if the excitation is m the blue range the emission is expected m the green So if one wants to use many different probes which emit different spectra it is evident that they must be close m wavelength, and also often overlap,

as a consequence, spectral resolution is of critical importance to be able to cUscriminate between the different probes.

According to tine method of die present invention, individual probes (a probe as refeπed to herein in this document also refers to a combinatorial probe) are assigned a pseudocolor (i.e., by an RGB algorithm) or an artificial color (i.e., a predeteπnined color according to a classification algorithm) and the information is displayed on a computer screen. The use of multicolor fluorescence opens up a possibility of extending FISH into important clinical applications which may benefit from multiple probes. Examples include aneuploidy and chromosome structural studies, detection of marker chromosomes and complete FISH karyotypes. Since multiple information may be gained from a single hybridization, throughput is increased and internal standards may be used in order to assess gene dosage effects or to determine die extent of deletions.

The metiiod of the present invention, utilizes detection of fluorescence excited by a white or coherent monochromatic light source m few nanow spectral bands and a sensor with cooled CCD. Thus, multiple spectra, each representing a different probe, may be simultaneously measured. This, m turn, increases the speed and accuracy of image acquisition, compared to conventional approaches which take multiple snapshots of chromosomes and then reconstruct the image, a process which is time consuming and generates artifactual results, all as described above. Hence, the present invention represents a highly significant progress over the state-of-the-art cytogenetic imaging, because it allows more sensitive, rapid and accurate detection of multiple probes.

Thus, according to the present mvention there is provided a fluorescent in situ hybridization method which method includes die steps of (a) providing a ceU nuclei having chromosomes, the chromosomes being hybridized witii at least one nucleic acid probe, each of the at least one nucleic acid probe including at least one nucleic acid molecule, each of the at least one nucleic acid molecule being labeled witii at least one fluorophore; (b) viewing the cell nuclei through a fluorescence microscope, the fluorescence microscope being optically connected to an imaging spectrometer, the fluorescence microscope and die imaging spectrometer being for obtaining a spectrum of each pixel of the ceU nuclei by (i) collecting mcident light simultaneously from all pixels of the cell nuclei using collimating optics; (n) passing the incident collimated hght through an interferometer system having a number of elements, so that the hght is first spht into two coherent beams which travel in different directions inside die interferometer and tiien the two coherent beams recombine to interfere with each other to form an exiting light beam; (iii) passing the exiting hght beam through a

focusing optical system which focuses the exiting hght beam on a detector having a two-dimensional aπay of detector elements, so that at each instant each of the detector elements is the image of one and always die same pixel of the cell nuclei for the entire duration of the measurement and so tiiat the real image of the cell nuclei is stationary on the plane of the detector aπay and at any time during the measurement the image is still visible and recognizable, and so tiiat each of the detector elements produces a signal which is a particular linear combination of hght intensity emitted by die pixel at different wavelengths, wherein the linear combination is a function of die instantaneous optical path difference; (iv) rotating or translating (i.e., scanning) one or more of the elements of the interferometer system, so that the optical patii difference between the two coherent beams generated by the interferometer system is scanned simultaneously for all the pixels of the cell nuclei; and (v) recording signals of each of die detector elements as function of time using a recording device to form a first spectral cube of data; and (c) interpreting die first spectral cube of data using a mathematical algorithm.

The nucleic acid probes may include loci, fragmented chromosomes, yeast artificial chromosomes each including an insert, plasmids, cosmids, phagemids or viral vectors each including an insert, complete (i.e., whole) genomes of a species or a cancerous tissue and combinations thereof. The fluorophores may be single fluorescent dye or a combinatorial fluorescent dye which are various combinations of single fluorescent dyes. The cell nuclei may be a nuclei during interphase, a nuclei during mitosis and a nuclei during meiosis and accordingly die chromosomes may be interphase chromosomes, chromosomes during mitosis or chromosomes during meiosis. The number of nucleic acid probes may be one, two, three, for, five, six, seven, eight, nine, ten, eleven, twelve, thirteen, fourteen, fifteen, sixteen, seventeen, eighteen, nineteen, twenty, twenty one, twenty two, twenty three, twenty four or greater than twenty four, each of the probes include a different fluorophore or a different combination of fluorophores (i.e., a combinatorial probe). The mathematical algorithm may be: (a) a Red-Green-Blue color image computation using predefined wavelength ranges; (b) a classification mapping analysis computing for the spectrum of each of the pixels a spectral difference from at least one reference spectrum; (c) a linear combination analysis, the analysis is for a background subtraction combined witii classification mapping analysis; (d) a principal component analysis; and (e) any other algorithm suitable for pixels classification according to tiieir associated spectra.

Reference is now made to the following examples, which together with the above descriptions, dlustrate the invention.

EXAMPLE 1 :

IMPROVED FLUORESCENT IN SITU HYBRIDIZATION fFISHI USING SPECTRACUBE™ A LINEAR COMBINATION ALGORITHM AND A CLASSIFICATION MAPPING ALGORITHM

Spectral bio-imaging using the SpectraCube™ system combined witii die method of the present invention enhances the usefulness of FISH by allowing the simultaneous detection of a large number of probes, in a single measurement and witii high accuracy. As a consequence, the efficiency and die rehabdity of detection of genetic abnormalities by FISH are greatly increased.

As detailed above, fluorescent in situ hybridization (FISH) plays an increasingly important role in many research and diagnostic areas. Since its first introduction in the 70 's the FISH technique has made significant progress, enabling the detection and identification of single gene sequences, partial chromosome sequences and even whole chromosomes (i.e., chromosome painting). The many applications of FISH range from early detection of diseases, to prenatal diagnosis, aneusomy and others, to discover and tiiereafter treat genetic diseases and abnormalities. Due to the high sensitivity and selectivity of FISH, which is based on hybridization of homologous nucleic acid sequences, even short sequences as small as 1 kilobase (kb) can be observed (and this will probably improve with time to enable the detection of sequences as short as 15-30 base pairs and, as a consequence, of point mutations). FISH can be applied both to interphase and metaphase cells and even to whole tissues, enabling a broad range of applications both in the fields of cytogenetics and pathology. FISH is improving hand m hand with the improvements of DNA probes, fluorescent dyes (especially die introduction of combinatorial probes), fluorescence microscopy, high performance CCD cameras and imaging techniques. The ability to detect many probes simultaneously has already been shown in die literature to make FISH an efficient diagnostic tool [Rudkin and Stollar ( 1977) Nature 55, 172-173]. However, the existing methods are cumbersome and difficult to use. As wtil be exemplified hereinbelow, the detection of many probes is greatly improved by die SpectraCube™ system combined witii appropriate algorithms, because of its spectral resolution and sensitivity. To illustrate this capability, the reader is now refeπed to Figures 5a-c, which include an example of an interphase FISH measurement performed witii chromosome 1 and chromosome 17 specific DNA probes tagged with the fluorophores Texas-Red and Rhodamine,

respectively, whose fluorescence spectra are very similar. The chromosome 1 probe was a midsatellite probe for the subtelomeric region of the chromosome and was tagged witii Texas-Red linked to die DNA probe via biotin post hybridization. The chromosome 17 probe was an a satellite probe for the centromeric region of the chromosome and was tagged with Rhodamme, linked to die second DNA probe via digoxigenin post hybridization. Figure 5a shows the original image, the way it looks to the eye through the microscope; Figure 5b shows the same sample, after being measured and processed by die SpectraCube™ system; and, Figure 5c shows the fluorescence spectra of the Texas-Red (marked as T) and Rhodamine (marked as R) fluorophores.

As seen n Figure 5c, the spectral peaks of Texas-Red and Rhodamme differ merely by 15 nm, and tiierefore it would be very difficult to distinguish between them using a filter-based system.

Looking at a color FISH image through a microscope as shown in Figure 5a, the confidence level of recognizing the coπect number of dots (marked 1-4) and of probe types appearing in the image is not particularly high. As shown in Figure 5b, the SpectraCube™ system, on the other hand, taking advantage of the spectrum measured for each pixel, is able botii to verify the existence of the dots, to count them exactly, and to discriminate between the different pairs with a high level of confidence, due to the small spectral difference between them. By artificial coloring of Texas-Red and Rhodamine fluorescence, as shown m Figure 5c the location of probe specific fluorescence could be determined with high accuracy wherein dots 1 and 2 are of Texas-Red and dots 3 and 4 are of Rhodamine. Figures 6a-b are an example of FISH measurement after hybridization of a nuclear DNA in interphase with six different probes. Figure 6a shows the original image; Figure 6b shows the SpectraCube™ measurement, spectral processing and artificial color display of all die detected pairs; and, Figure 6c the spectra of the six chromophores after hybridization (marked according to die chromosomes each of which labels: 1, 8, 10, 11, 17 and X), as detected through a triple dichroic filter using the SpectraCube™ system. (For details regarding flourophores, probes and chromosomes the reader is refeπed to the following description, Table 2 below and to Chroma Corp. Cat. No. 61502.

It is apparent from Figure 6a, showing the original RGB image of the interphasic cell nucleus, that it is difficult to distinguish the colors from one another by eye or even by using a simple RGB color measurement. An experienced observer may, in the best case, detect tiiree different colors of the six. Figure 6b, however, shows the same sample shown in Figure 6a, after processing

the spectral data with propnetary classification algonthms for background subtraction and classification (see, details above), and die resulting dots have been highlighted with artificial colors as follows brown - BI; cyan - C, blue - B2; yellow - Y, green - G, and red - R, while die background was given a black - B3, artificial color As observed, it is possible to see all die six pans of fluorophores and to easily differentiate among die paus

It should be further noted tiiat one pair, the one highlighted m blue (B2), can hardly be noticed by eye, or by usmg a color camera, however, it is detected after applymg a background subtraction algonthm on the spectral cube (compare Figure 6 a with 6b).

The probes used were five satelhte probes for the centromeπc regions of chromosomes 8, 10, 11, 17 and X. and a midsatellite probe for the subtelomenc region of chromosome 1 The fluorophores used to label each of die above chromosomes and die DAPI counter stain (backg ), theu emission peak and artificial displayed color classification are summanzed m Table 2

From the normalized spectral signatures of each of the six fluorophores shown m Figure 6c, it is clear that a system based on filters measuring at a few relatively wide spectral ranges, is not able to differentiate reliably between the different probes species, because of the large overlap between the spectra Such a system is more dependent on die absolute measurement of the mtensity of each probe, and tiierefore it is more affected by background signals and noise It should be further noted tiiat spectral overlappmg sometimes occurs also with auto- fluorescence originating from the cell itself In this case too, the avadabihty of spectral information for each pixel enables the elimination of the auto-fluorescence contribution, and yields more accurate results

con uga e v a an - igox gemn an o y pre y ze igoxigenin containing probes ;

•~)

⁰ fluorescein-5-iso-thiocyanate, conjugated via anti-biotin antibody to pre hybridized biotin containing probes; 4 4',6-diamidino-2-phenylindole used for counter staining.

Having measured die full spectrum of each point on the image, may also help overcome specificity problems of the probes. In fact in some cases, a probe that matches a certain chromosome DNA sequence, has also a lower specificity to a different (usually similar) sequence, and it hybridizes witii a lower probability to the second sequence too. This leads to the spurious appearance of too many probes of a certain type. However, the fluorescence spectrum m the second case is very slightly shifted witii respect to the first one, due to a small change m the chemical environment of the probe. The SpectraCube™ system, thanks to its spectral resolution and sensitivity, may eliminate this artifact. A similar artifact exists for probes which are not washed out during sample preparation, and contribute to false positive diagnosis. The SpectraCube™ system combined with the method of the present mvention, therefore, helps lowering the risk of wrong diagnosis.

Generalizing to a large number or similar dyes, the examples of Figures 5a-b and 6a-c show tiiat it is possible to detect and distinguish a large number of probes, and, provided tiiere are small spectral differences between them, the SpectraCube™ will detect and identify them in one measurement.

It is clear to one ordinarily skilled in the art that otiier and/or additional known and yet to be discovered or developed fluorophores and fluorophores combinations may be used in various FISH applications as detailed above to detect large number of loci simultaneously, to paint each chromosome of a karyotype in a distinguished color, etc. A list of flourophores used m state of the art cellular and molecular biology may be found in Kasten (1993) Introduction to fluorescent probes: Properties history and applications, m Fluorescent and luminescent probes for biological research, Mason Ed. Academic Press Limited, London, pp. 24-31. It is also clear to one ordinarily skilled in the art that other labeling techniques such as for example bioluminescent and chemoluminescent and also non-fluorescent labeling strategies may be similarly applied.

Thus, using the SpectraCube™ system for FISH analysis enjoys a major advantage as foUows. The SpectraCube™ system, due to its high spectral resolution, enables simultaneous detection of numerous probes, whereas using conventional means to perform FISH (e.g., using a fluorescence microscope) limits the number of probes to be used in a single hybridization to two - four probes. Therefore, employing the SpectraCube™ system for FISH analyses save effort and time. Furthermore, while employing the SpectraCube™ system for FISH analysis

a smaller number of cells are required for full analysis, an important feature in cases where the number of cells to be analyzed is limited.

EXAMPLE 2:

SIMULTANEOUS VISUALIZATION OF ALL HUMAN CHROMOSOMES IN DIFFERENT COLORS USING FLUORESCENT IN SITU HYBRIDIZATION AND SPECTRAL BIO-IMAGING.

The emergence of multicolor FISH has broadened the applications of molecular cytogenetics in basic research and genetic diagnosis. .All existing multicolor FISH techniques require the use of fluorescent probes whose emission spectra can be separated with optical filters [Ried et al., ( 1992) Simultaneous visualization of seven different DNA probes by in situ hybridization using combinatorial fluorescence and digital imaging microscopy. Proc. Natl. Acad. Sci. USA 89, 1388- 1392; and, Ried (Jan. 1994) Fluoreszenz in situ Hybridizierung in der genetischen Diagnostik, Faculty of theoretical medicine, Ruprecht-Karls University Heidelberg]. This requirement limits the number of dyes which can be distinguished m a given sample. According to die present mvention provided is a novel approach for FISH, employing the SpectraCube™ system and the method of the present invention to measure and analyze multiple spectrally overlapping labeled probes (single and combinatorial). In this Example, spectral bio-imaging which, as delineated above, is a combination of Fourier spectroscopy, CCD- imaging and optical microscopy enabling the measurement of definitive spectral data simultaneously at ad pomts of a biological sample, was used to visualize hybridization based multicolor bands along all (i.e., 24) types of human chromosomes and to generate a color map of the human karyotype.

For this purpose, 24 chromosome pamts (1 through 22, X and Y, Table 4) each labeled witii a different combmation of five or less different flourophores (a through e, Table 3), (see Table 3 for the different fluorophores and tiieir spectral characteristics and Table 4 for the assignment of the fluorophores listed in Table 3 to obtain the 24 chromosome paints), were simultaneously hybridized witii human mitotic chromosome spreads of male white blood ceds, prepared for hybridization essentially as described in Ried et al. [Ried et al., (1992) Simultaneous visualization of seven different DNA probes by in situ hybridization using combinatorial fluorescence and digital imaging microscopy. Proc. Natl. Acad. Sci. USA 89, 1388-1392]. Hybridized chromosomes were viewed tiirough an inverted

fluorescence microscope connected to die SpectraCube™ System and were analyzed.

TABLE 3:

from Amersham

TABLE 4:

With reference now to Figures 7a-e, 8a-b and 9a-b. Figures 7a-e show normalized spectra of 24 individual pixels, each of a different type of human chromosome ( 1-22, X and Y). Numbers 1-22 and letters X and Y, refer to the chromosome type of which each of the spectra presented were derived. Note tiiat the spectrum obtained from each of die 24 human chromosomes, as shown in Figures 7a-e, differ from all other spectra. This difference may be large (compare, for example, the Ca. 530 nm emission peak of chromosome 15 and 11 in Figure 7c) or small (compare, for example, the Ca. 530 nm emission peak of chromosome 22 and Y in Figure 7e) and, in some spectral ranges may even disappear (compare, for example, the Ca. 680 run emission peak of chromosome 22 and Y in Figure 7e). Nevertheless, as further shown in Figures 7a-e, even a minor difference between very similar spectra can be detected using the SpectraCube™ system and the method of the present invention. It is however clear from this description that the ability of the method of the present mvention to detect differences among spectra, to a large extent depends upon appropriate fluorophores and fluorophore combinations selected, yet, as will be appreciated by one ordinarily skilled in the art and even by one expert in the art, the ability herein demonstrated, far beyond exceeds tiiat of any prior art cytogenetic technique. Figure 8a shows an RGB image of thus described painted human chromosomes, whereas Figure 8b shows a color human karyotype derived from die painted chromosomes of Figure 8a. Since it is not possible to literally describe 24 different colors, colored Figures 9a and 9b which are otherwise identical to black and white Figures 8a and 8b, respectively, are also enclosed. Note tiiat each of the chromosome pairs is painted in a different color and tiiat the color karyotype (Figure 9b) is readily denved from die color image (Figure 9a).

The algorithm used to obtain and display die image of Figures 8a and 9a was an RGB algorithm as described above and as exemplified in Figure 4, wherein R = 640-740 nm; G = 550-640; and, B = 530-550 nm. However to obtain a more unified image in terms of intensities, a special modification of the RGB values obtained was exercised. This modification, known in the color imaging art as "contrast stretching", [see for example, ENVI™ User's guide, The environment for visualizing images Version 1.1 July 1994 Edition, BSC limited Liabdity Company] includes (a) deterrnining die distribution of each of the RGB values; (b) defining a look-up table for maximizing differences among intensities within the determined distribution; and (c) displaying the modified RGB image, now having a maximal color variation for each original different spectrum. This simple modification of the original RGB image is actually a limited version of a classification algorithm

as described above. It is however clear to one ordinarily skilled in the art that other algorithms may equivalently or better suit the purpose of displaying similar images. As detailed and exemplified hereinabove (e.g., Example 1), classification mapping will enable to present any of the human chromosomes in any predetermined artificial color to obtain an even more chromosome distinctive color patterns. As further detatied above, a principal component analysis may also be found suitable, wherein each meaningful component or combinations thereof will be attributed a different predetermined artificial color. Yet, additional algorithms capable of differentiating similar spectra and attributing a different predetermined artificial color (or pseudo color) to pixels having a different spectrum may also be found suitable for color karyotyping according to the method of die present invention.

In this Example, the use of 24 different single and combinatorial probes combined from five different basic fluorophores (a through e, Table 3) was demonstrated for human color chromosome karyotyping. Nevertheless, some other species have a greater number of chromosomes, which perhaps requires the use of more complicated combinatonal probes combined of more basic fluorophores. Yet, it should be noted that chromosomes, including human chromosomes, can also be classified to size groups, which, for some applications minimize the need for as many different colors since chromosomes belonging to different size groups may be similarly colored yet easily recognized according to their relative size. This could be achieved by manual inspection, or alternatively using any morphological algorithm.

EXAMPLE 3:

DETECTION OF MULTIPLE CHROMOSOME TRANSLOCATIONS IN BREAST CANCER CELLS

As demonstrated, die metiiod of die present invention can provide a complete color karyotype of normal blood cells. In many cases conventional karyotyping (e.g., using G-banding or R-banding techniques) is used to detect chromosomal abeπations such as translocations associated witii genetic disorders (e.g., 21q22 trisomy in Down's syndrome, chromosome 18 (or a fragment thereof) trisomy and chromosome 13 (or a fragment thereof) trisomy) or mahgnancies (e.g., a translocation between the distal ends of chromosomes 9 and 22 in leukocytes from patients with chronic myelogenous leukemia and, a chromosomes 8 and 14 translocation in lymphocytes of patients with Burkitt's lymphoma). In this

Example the capabilities of the SpectraCube™ system combined witii die metiiod of die present invention to detect multiple chromosome translocations in breast cancer cells is demonstrated.

Witii reference now to Figures lOa-b, l la-b 12 and 13. Chromosome spreads of breast cancer cells were hybridized witii the 24 chromosome paints (1 through 22, X and Y) as detailed in tables 3 and 4 above and were spectrally imaged as detailed under Example 2 above. Figures 10a and 11a show DAPI R- banding of a chromosome spread as was photographed using a conventional fluorescence microscope. It wdl be appreciated that although the resulting karyotype is abnormal to a large extent, it is impossible to identify specific translocations of chromosomes in Figures 10a and 1 1a. Figures 10b and l ib show (in black and white and color, respectively) an RGB image of the same spread as was obtained using die SpectraCube™ system and die method of the present invention. When compared with Figures 9a-b, presenting a normal human karyotype denved under otherwise identical experimental conditions, it is apparent that manv of the abenations containhi"; chromosomes shown in Figures 10b and l ib contain parts of various normal human chromosomes. The translocated chromosomes (right) of Figures 10b and l ib, along witii the R-banded chromosomes (left) are shown in Figures 12 and 13. Note tiiat some of the translocated chromosomes shown in Figures 12 and 13 include fragments originated from two, three and even four different chromosomes. Smce specific chromosome translocations and otiier chromosomal abenations (e.g., gene amplification) were previously associated with early or stage specific mahgnancies and, since the metiiod of the present invention tremendously increases the ability to detect and characterize such translocations and otiier abenations. the ability of early detection and stage classification of malignancies exercising the method of the present invention will be benefited. Furthermore, using the metiiod of the present invention wdl enable to establish yet new chromosome specific abenations (e.g., translocations) recunently associated witii specific malignancies and eventually to provide a comprehensive guide for such translocations. In addition, such recuπent abenations may lead to the isolation of genes that by being activated or alternatively inactivated are associated witii die malignant processes.

In this context it is important to notice that the fluorophores employed in this and die former Example 2 (as listed in Table 3 and shown in Figures 7a-c) collectively emit in the 480-730 nm range. Thus DAPI can be simultaneously used for counter staining since its emission is in the blue, well below this spectral range. Thus, as shown in Figures lOa-b, 1 la-b, 12 and 13, it is possible to simultaneously observe the very same chromosome spreads using the conventional monochromatic

R-banding approach and die multi-color approach of the present invention. It is clear that the SpectraCube™ system is capable of providing a conventional DAPI R-banding image by limiting the examined spectral range to blue. Thus, for a comparative purpose a single chromosome spread may be viewed using the SpectraCube™ system and the method of die present invention as a DAPI R- banded karyotype as well as a color karyotype. Hence, when chromosome translocation events are studied accordmg to the method of the present invention the DAPI R-banded karyotype can provide additional information and to precisely point out which region(s) (i.e., bands) of any specific chromosome are mvolved in a specific translocation event.

EXAMPLE 4:

ADDITIONAL FISH APPLICATIONS

From the above descriptions it is clear that ( 1) several types of probes may be used for FISH, these include loci specific probes, chromosome paints and whole genomes; (2) the analyzed chromosomes may be during interphase, mitosis or meiosis; and, (3) dozens of probes of all types may be simultaneously hybridized to the chromosomes and, provided that each of the probes has a somewhat different spectrum, the SpectraCube ™ system as used accordmg to die method of the present mvention can spectrally detect each of die probes and present its spatial aπangement by attributing pixels presenting each of the spectra an RGB color (i.e., pseudocolor) or a predetermined artificial color. Thus, for example if the method of the present invention is to be used for mapping a newly isolated gene(s) (or other DNA sequences), a single procedure may be employed to map die gene(s) to tiieir chromosomal bands. Exemplified for two new genes, to this end 26 different probes may be prepared as follows: 24 chromosome paints and two loci specific probes (i.e., the newly isolated genes fluorescently labeled). The probes are then mixed and simultaneously hybridized preferably to mitotic chromosomes which are also DAPI counter stained. The result is a 24 (for a male, or 23 for a female) color karyotype, similar to the one presented in Figure 9b, on which two loci specific signals (dots attributed to die loci specific probes) in yet two different colors point out the chromosome locations of the newly isolated genes which are then associated with a specific chromosomal band by generating an R-banded image as explained above.

In many cases, few loci specific probes are mapped to a single chromosomal band, yet which is distal and which is proximal is not established.

Usmg tiie metiiod of the present mvention to simultaneously detect each of the few probes as each appears in a different RGB or artificial color, will, m many cases, enable to determine the relative aπangement of closely mapped sequences

The SpectraCube™ system and die metiiod of the present mvention may also be used to detect interphase chromosome three dimensional anangements The reader is refeπed again to Figures 8a and 9a On the upper nght corner presented is a nuclei during interphase (marked NI) hybπdized with the chromosome pamts listed m Table 4 above Examination of die color pattern of this nuclei reveals a unique feature Note for example that both the chromosome 2 pair (m red) are located m me lower part of the nuclei and tiiat the chromosome 6 pair (m purple) are both located m the opposite pole Little is so far known about the chromosome organization duπng interphase, yet it is reasonable to suspect that changes occur m the chromosome organization duπng interphase m malignant cells Thus, the metiiod of the present mvention may be of great value for early detection of vaπous mahgnancies, defining the stage of a malignant disease, and hence better adjust a treatment to examined patients, etc It should be noted tiiat usmg the SpectraCube™ system combmed witii die metiiod of the present mvention and a three dimensional reconstruction means (e.g , a confocal microscope) may be used to extract three dimensional information of chromosome organization during interphase

Many cancers and genetic disordeis are characteπzed by chromosome deletions, translocations and other reaπangements and gross abnormalities (e g , gene amplification) As demonstrated m Example 3 above, usmg the method of die present mvention will enhance the ability to detect such abnormalities Furthermore, it is clear that die metiiod of die present mvention is highly suitable for comparative genomic hybndization (CGH ) and for reverse chromosome painting as descnbed above

One of the common chromosomal abenations is associated witii Down's- syndrome It was long ago established tiiat Down's syndrome results due to tnsomy of chromosome 21 More careful exammation revealed that a specific region of chromosome 21 (21q22) is always associated (l e , appears m tnsomy) with this common syndrome However, m some cases the karyotype of individuals affected with Down's syndrome is apparently normal as determined by conventional G- or R-bandmg karyotyping techniques The widely accepted explanation to this phenomenon is that m these cases die tnsomy is of a fragment denved from die 21q22 chromosome region which fragment is smaU and below the resolution of the conventional banding techniques However, usmg the SpectraCube™ system combmed with the method of the present invention will

enable to detect tiiese so far undetectable chromosome 21 tisomies in embryonic cells obtained for example via chorionic villi sampling and to enable a more educated genetic counseling to high risk women. It should be noted tiiat chromosome 13 and chromosome 18 or fragments thereof were also reported to appear in trisomies resulting in birth of strikingly abnormal children and tiiat the metiiod of die present invention can be similarly applied for a prenatal diagnosis of these devastating chromosome 13 or 18 trisomies.

The method of the present invention, combined witii the rapidly developing techniques of separating embryonic cells from peripheral blood of a pregnant woman will be of great value for low-risk prenatal karyotyping for the detection of chromosome 21 trisomies and other, less frequent chromosome abnormalities.

Using the SpectraCube™ system and the method of the present invention combined witii chromosome telomeres specific probes, each of the telomers (48 in human males, 46 in females) appears in a different color, will enable a comparative study of all telomeres m an examined species.

In the study of evolutionary related species and in the study of model systems (for example mouse as a model system for human) it is in many cases required to obtain comparative genome maps in which chromosomes of two or more species are aligned according to their sequence similarities and tihus their chromosome-borne genetic information. Using the metiiod of the present invention will facilitate obtaining such comparative maps. Consider for example the preparation of a human-mouse chromosome comparative map. For this purpose a complete set of chromosome paints of one of the species (e.g., human) are to be simultaneously hybridized with chromosome spreads of die otiier species (mouse in the given example) and analyzed as described above. The result is an image of the mouse karyotype painted with the human chromosome paints. Thus, an alignment can be made between the karyotypes of the two species.

Many other applications for FISH were so far described in the arts literature. One example is in the study of gene expression wherein by usmg loci specific probes hybridized witii interphase nuclei obtained at intervals from a synchronized cell culture one can determine their order of replication (i.e., replicated genes appear ad four dots and non-replicated genes appear as two dots), wherein, as a rule of thumb, early replicating genes are expressed in the examined cells and late replicating genes are not. The method of the present invention is highly suitable for this type of analysis since dozens of probes each having a slightly different spectrum can be analyzed simultaneously in a single hybridization followed by a single imaging step to detect mem all. In fact the

method of the present invention can be used for any FISH apphcation so far described or yet to be described.

Whde the invention has been described witii respect to a limited number of embodiments, it wdl be appreciated tiiat many variations, modifications and otiier applications of the invention may be made.