The invention relates to a laser scanning microscope arrangement and a method to operate a laser scanning microscope arrangement.

In non-linear optical microscopy, confocal microscopy or fluorescent microscopy, laser pulses are commonly used to excite a physical sample which in return emits light of a different wavelength. The emitted light can be measured and used to form a visual representation of the physical structure of the sample. Two-photon excitation

microscopy, for example, is a commonly used fluorescence imaging technique that allows visual imaging of living tissue up to about one millimeter in depth.

Time-varying background light poses a serious problem for such laser scanning microscopy, as it overlays the emitted light from the sample and makes the imaging process harder for users. This is particularly problematic if the time-varying background light varies on time scales that are longer than the time period of the excitation laser pulse, such as in the case of AC lighting. In order to adjust for the background light, laser scanning microscopy is preferably performed in darkened environments.

According to the invention, this problem is solved by a laser scanning microscope and a method to operate a laser scanning microscope according to the independent claims. According to the invention, the laser scanning microscope arrangement comprises a pulse laser oscillator unit configured to create a laser beam, which is triggered by a laser pulse clock signal oscillating at an excitation frequency f _osc in order to excite a physical sample. The sample can be, for example, tissue treated with fluorescent agents or any other type of physical sample emitting light of different wave length upon stimulation by a laser pulse of specified frequency.

A laser beam shaping means might be provided after the pulse laser oscillator unit, as well as an X-/Y-scanning means in order to properly direct the shaped laser beam to the sample.

An optical signal detection unit is used to detect the light emitted from the sample and output an electric signal in response. The electrical signal might be indicative of the strength, frequency or other characteristics of the emitted light. The electric signal might be an analog signal. The optical signal detection unit might comprise at least one optical detector such as a photodetector. The optical signal detection unit might comprise a multitude of optical detectors in order to detect different characteristics of the emitted light at the same time. The optical signal detection unit therefore might output a multitude of electrical signals indicative of different characteristics of the emitted light at the same time.

Beam shaping means, scanning means and optical signal detection unit might be provided in a common microscope arrangement.

An electrical signal acquisition unit is used to acquire and sample the electric signal output from the optical signal detection unit by a sample clock signal oscillating at a sampling frequency f _s . The sampling frequency f _s is higher than the excitation frequency fosc- In typical embodiments, the sampling frequency can be 5 to 1000 times higher than the excitation frequency, preferably 10 to 100 times. Particularly for ratios up to 100, the sampling frequency can be an integer multiple X of the excitation frequency, while for higher ratios, the sampling frequency can also be a non-integer multiple of the excitation frequency. The signal acquisition unit can be realized as an integrated circuit or as a program module in a personal computer equipped with an electrical interface module. The resulting sample data is stored in a memory unit, such as a hard drive, a semiconductor flash memory, or any electronic storage means.

According to the invention, a signal conditioning unit is provided, which is configured to receive past sample data from the memory unit and determine, preferably by statistical analysis, background noise values. The signal conditioning unit can be realized as an integrated circuit or as a program module in a personal computer.

According to the invention, the signal conditioning unit passes the sample data and the determined background noise values to a noise reduction unit, which is configured to compensate the signal data for background noise by subtracting the background noise values from the signal data values. The noise reduction unit can be realized as an integrated circuit or as a program module in a personal computer.

This enables to highly improve the results of the microscope measurements. It particularly enables to perform measurements irrespective of ambient light, in particular time-varying ambient light.

According to a further embodiment of the invention, the signal conditioning unit is configured to set up a number of X datasets D, with i = 0 ... X-1 , each comprising sample data values at consecutive delay times t, relative to the laser pulse clock signal. Each dataset represents measurements at a given delay time after the laser pulse. Usually, the emitted signal values will be measured at a certain delay time after the laser pulse, which might depend on the sample measured. In this embodiment, the signal conditioning unit is configured to determine a dataset D _s and delay time t _s comprising the emitted signal values and determine the average background noise values by averaging over the values in the remaining datasets.

The signal conditioning unit, in order to determine the dataset D _s , might be configured to calculate an average value or a spread for each dataset D,, compare the average values or the spreads with each other, and select D _s as the dataset with highest average value or highest spread. This is because the emitted signal values will typically have a larger average and spread than the background noise values. In a further embodiment, the signal conditioning unit, in order to determine the dataset D _s , is configured to transmit the datasets D, to a user interface, which is configured to present the datasets to the user, receive a manual selection, and inform the signal conditioning unit of the manually selected dataset Ds. In this embodiment, an informed user makes the decision on which dataset to use for extracting the emitted signal values, and which datasets to use for calculating the background noise values. The user interface might comprise an electronic interface, a screen and an electronic input means such as a keyboard.

In a further embodiment, the signal conditioning unit is configured to compare the sample data values to a predetermined threshold value, which is preferably retrieved from the memory unit or based on statistical analysis, classify the sample data values as signal values or noise values, and calculate the background noise value as the average value of the noise values.

In this embodiment, the signal conditioning unit might be configured to calculate a moving average of the classified noise values within an averaging window prior to or around the classified signal values. The averaging window size might be larger or equal than the time period 1/fosc of the laser pulse signal, preferably an integer multiple of 1/fosc, preferably a multiple of 1 , 2, or 3.The signal conditioning unit might select a number of identified noise values prior to or in the vicinity of a signal value to calculate the average background noise value. In particular, 5 - 10 noise values prior and after an identified emitted signal value might be considered to calculate the average value.

The resulting compensated data might be passed to a visualization unit which calculates, out of the compensated sample data values, an image of the sample to be shown to the user. The visualization unit might also compensate for scanning

distortions. It might be realized as an electronic interface and a screen. The invention further relates to a method to operate a laser scanning microscope arrangement, comprising the steps of:

creating, by a pulse laser oscillator unit, a laser beam, triggered by a laser pulse clock signal oscillating at an excitation frequency , in order to excite a sample;

detecting, by a optical signal detection unit, predetermined characteristics of light emitted from the sample and outputting, by the optical signal detection unit, an electric signal in response;

sampling, by an electrical signal acquisition unit, the electric signal output from the optical signal detection unit by a sample clock signal oscillating at a sampling frequency f _s , wherein f _s is higher than f _osc ;

storing, by a memory unit, the discrete electrical sample data values of the signal acquisition unit.

According to the invention, the method further comprises the steps of

retrieving, by a signal conditioning unit, past sample data values from the memory unit and determining, by the signal conditioning unit, preferably by statistical analysis, background noise values;

subtracting, by a noise reduction unit, the background noise values from the sample data values.

It might be provided that the sampling frequency f _s is an integer multiple X of the excitation frequency , X preferably being a value of 5 up to 1000, preferably 10 up to 100.

It might be provided that in step e), the signal conditioning unit sets up a number of X datasets D, with i = 0 ... X-1 , each comprising sample data values at consecutive delay times ti relative to the laser pulse clock signal; determines a dataset D _s and delay time t _s comprising emitted signal values; and determines the average background noise values by averaging over the values in the remaining datasets.

It might be provided that the signal conditioning unit, in order to determine the dataset D _s , calculates an average value or a spread for each dataset D,, compares the average values or the spreads with each other, and selects D _s as the dataset with highest average value or highest spread. In an embodiment, the signal conditioning unit, in order to determine the dataset Ds, might transmit the datasets D, to a user interface, which presents the datasets to the user, receives a manual selection, and informs the signal conditioning unit of the manually selected dataset D _s .

In step e), the signal conditioning unit might compare the sample data values to a predetermined threshold value, which is preferably retrieved from the memory unit or based on statistical analysis, classify the sample data values as signal values or noise values, and calculate the background noise value as the average value of the noise values. In particular, the signal conditioning unit might calculate a moving average of the classified noise values within an averaging window prior to or around the classified signal values. The averaging window size might be larger or equal than the time period 1 /fosc of the laser pulse signal, preferably an integer multiple of 1 / , preferably a multiple of 1 , 2, or 3.

Further features of the invention will become apparent from the following claims, figures, and description of embodiments.

An exemplary embodiment of the invention is described in the following figures:

Fig. 1 shows a schematic block diagram of an exemplary embodiment of a laser scanning microscope arrangement according to the invention;

Fig. 2 shows a schematic line chart of the pulses and measurements occurring in an embodiment of the invention;

Fig. 3 shows a schematic representation of the datasets D, created by the signal conditioning unit in an embodiment of the invention.

Fig. 1 shows a schematic block diagram of an exemplary embodiment of a laser scanning microscope arrangement according to the invention. The arrangement comprises a pulse laser oscillator unit 1 for oscillating a pulsed laser beam, a beam shaping unit 2 such as a dichroic mirror, and a scanning unit 3 such as a galvanometer mirror for scanning the laser beam in X and Y directions. The pulsed laser beam is directed, preferably through a not-shown objective lens, onto a sample 4 which comprises fluorescent material. The arrangement further comprises an optical signal detection unit 5 such as a photodetector or a photomultiplier to convert the emitted light signal into an electrical signal. The optical signal detection unit might comprise a multitude of photodetectors or photomultipliers in order to convert a range of different wavelengths of the emitted signal into different electrical signals. The beam shaping unit 2, scanning unit 3, and optical signal detection unit 5 might be comprised within an optical laser scanning microscope 12. A control unit 1 1 is provided which receives the oscillating laser pulse clock signal at a frequency of f _osc and the oscillating sample clock signal at a frequency of f _s . The sample clock has a higher frequency f _s than the laser pulse clock’s frequency f _osc . In certain embodiments the value of f _s might be 100 MHz and the value of f _osc might be 10 MHz. The control unit further provides the scanning unit 3 with X- and Y- scanning data.

The optical signal detection unit 5 outputs an analog electrical signal indicative of the characteristics of the received emitted signal, such as the strength of the emitted light. This electrical signal enters an electrical signal acquisition unit 6, which receives the sample clock signal from the control unit 1 1. The signal acquisition unit 6 samples the analog electrical signal according to the sample clock and stores the resulting discrete sample data values in an electronic memory 7, such as a semiconductor memory.

A signal conditioning unit 8 is provided, which retrieves past sample data values from the memory unit 7, and determines background noise values from past sample data.

For this, the signal conditioning unit 8 receives the sampling clock signal f _s and the laser pulse clock signal f _osc from the control unit 1 1.

In an embodiment of the invention, the signal conditioning unit 8 compares the sample data values to a threshold value, which is preferably retrieved from the memory unit 7, and classifies the past sample data values as signal values or noise values. It then calculates the average value based on the noise values alone. The threshold value can also be calculated based on statistical analysis. In an embodiment of the invention, the signal conditioning unit 8 calculates a moving average of the past sample data within an averaging window. It might calculate the moving average based on noise values alone, if these have been identified. The averaging window might be placed prior to identified signal values or around identified signal values. For example, in certain embodiments the averaging window might encompass ten noise values prior to one signal value. In other embodiments, the averaging window might encompass five noise values prior to one signal value and five noise values after that signal value.

In a further embodiment of the invention, the signal conditioning unit 8 sets up a number of X datasets D, with i = 0 ... X-1 , each comprising sample data values at consecutive delay times t, relative to the laser pulse clock signal. The values of the index i are indicated in Fig. 2, and Fig. 3 shows the created data sets Di for three laser pulse trigger signals. The signal conditioning unit 8 determines the dataset D _s and delay time t _s comprising the emitted signal values and then determines the average background noise values by averaging over the values in the remaining datasets.

In order to determine the relevant signal dataset D _s , the signal conditioning unit 8 calculates an average value or a spread for each dataset D,. The spread of the noise signals in the datasets D ₀ - D ₄ and D ₆ - D _g is indicated in Fig. 3. The signal conditioning unit 8 then compares the average values or the spreads with each other, and identifies the data set D _s as the dataset with highest average value or highest spread.

In a further embodiment, the signal conditioning unit 8, in order to determine the dataset Ds, transmits the datasets D, to a user interface 13, which presents the datasets to a user, receives a manual selection from the use, and informs the signal conditioning unit 8 of the manually selected dataset Ds.

The sample data and the calculated background noise value are passed from the signal conditioning unit 8 to a noise reduction unit 9. The noise reduction unit 9 subtracts the calculated average noise value from the sample data values. The resulting compensated sample data values are passed to a visualization unit which receives the X-/Y-scanning data from the control unit 1 1 in order to visualize the compensated sample data on a screen.

Fig. 2 shows a schematic line chart of the timing of the pulses and measurements occurring in an embodiment of the invention. The laser pulse clock with the frequency fosc is shown in dotted lines. It serves as a trigger signal and gives rise to laser pulses which are drawn in a bold line. The dotted bold line shows the emitted light signal from the sample incident on the optical detection unit 5.

Depending on the physical process underlying the signal generation from the sample (e.g. fluorescence, harmonic generation, coherent Raman scattering, etc.), the pulse form of the emitted signal and its relative timing might vary. Signals due to different physical processes can occur at different wavelengths, and therefore be detected on physically separated detectors and detected independently, or they can occur at the same (or similar) wavelength and therefore can be detected on a single detector. In the latter case, the signals may be separated in time or have a different temporal pulse form, and therefore they can be resolved separately by their different temporal signatures.

In order to measure the emitted signal, an electrical signal acquisition unit 6 samples the electric signal output of the optical signal detection unit 5. The sample clock is shown as a thin line. In this embodiment, the frequency of the sample clock f _s is ten times the frequency of the laser pulse clock f _osc . The measured sample data values are indicated in Fig. 2 as plus signs. It can be seen that most measurements represent background noise. In order to compensate for the background noise, the signal conditioning unit 8 analyzes past sample data values and calculates an average noise value.

For this, the signal conditioning unit 8 first discriminates between sample data values and noise values. This can be achieved by considering a threshold value, statistical analysis, or other methods, as described above. It might be known that for typical fluorescent processes, defined lag times occur between the laser pulse and the emitted signal, which allows the signal conditioning unit 8 to identify the signal values. It might also be known that typical signal values will exceed a certain signal threshold or are confined within a specific wavelength. When the signal conditioning unit has identified a number of noise values, it calculates an average noise values within an averaging window based on a moving average algorithm. The average noise value is shown in Fig. 2 as dash-dotted line. In this embodiment, the average noise value is calculated around each identified signal value and takes five leading as well as five lagging noise values into account for averaging. The calculated average noise value is then deducted from the sample data values in order to receive compensated sample data values.

Fig. 3 shows a schematic representation of the datasets D, created by the signal conditioning unit 8 in an embodiment of the invention as described above. By comparing the average values and/or the spread of the different datasets D,, the signal conditioning unit 8 can select the data set index i which relates to the time delay t, at which an emitted signal is received. In this embodiment, the index i=5, so that only the values in the data sets D ₀ - D ₄ and D ₆ - D _g are used to calculate the background noise values.

List of numerals

1 Pulse laser oscillator unit

2 Beam shaping unit

3 Scanning unit

4 Sample

5 Optical signal detection unit

6 Electrical signal acquisition unit

7 Memory unit

8 Signal conditioning unit

9 Noise reduction unit

10 Visualization unit

11 Control unit

12 Microscope

13 User interface