NORMALIZING THE RESPONSE OF A FLUORESCENCE INSTRUMENT

USING SPECTRAL RESPONSE

CROSS REFERENCE TO RELATED APPLICATION(S)

This application claims priority under applicable portions of 35 U.S.C. §1 19 to U.S. Patent Application Serial No. 62/052,132, filed September 18, 2014, and entitled: FLUORESCENT STANDARDS FOR AUTOMATIC RECALIBRATION, the entire contents of which are incorporated by reference herein. FIELD OF THE INVENTION

The invention relates generally to a method of calibrating diagnostic analyzers using fluorometry as a measurement mechanism.

BACKGROUND OF THE INVENTION

The present invention pertains to at least one clinical diagnostic analyzer conducting an immunoassay employing a fluorescence label. Typically, a fluorescence label is bound to antibodies or antigens having an affinity for the analyte of interest. The unknown analyte in the sample then binds with the labeled antibodies or antigens which are usually immobilized to a substrate. The unbound, labeled antibodies or antigens are subsequently washed away, and the concentration of bound, labeled antibodies or antigens is measured using fluorometry.

Fluorometry is the measurement of fluorescence. Fluorescence is the molecular adsorption of light energy at one wavelength and its nearly instantaneous re- emission at another, usually longer, wavelength. The instrument used to measure fluorescence is called a fluorometer. A fluorometer generates the wavelength of light required to excite the analyte of interest and then it measures the intensity of the resulting emitted light. The amount or quantity of emitted light is frequently proportional to the concentration of the analyte being measured. When employed in clinical diagnostic analyzers fluorometry provides extraordinary sensitivity, high specificity, simplicity, and low cost as compared to other analytical techniques. To insure the quality control of results from fluorometers, some form of stable reference standard is employed such as National Institute of Standards and Technology (NIST) SRM (Standard Reference Material) 2944 glass. SRM 2944 is a cuvette-shaped, bismuth-ion-doped glass, recommended for use for relative spectral correction of emission and day-to-day performance verification of steady-state flu- orescence spectrometers. Further information regarding SRM 2944 is described by Paul C. DeRose; Melody V. Smith; Jeffrey R. Anderson; Gary W. Kramer in the Journal of Luminescence, Volume 141 , pp. 9 - 14, entitled "Characterization of Standard Reference Material 2944, Bi-lon-Doped Glass, Spectral Correction Standard for Red Fluorescence" which is hereby incorporated by reference in its entirety.

One problem presented by fluorometers is that variations in the manufacture of clinical diagnostic analyzers are such that for a given fluorescence label; the population of clinical diagnostic analyzers will not provide the same analytical result for a specific quantity of analyte in a sample. These manufacturing variations result from differences in excitation light spectra from the laser diode, variances in transmission characteristics of optical filters, etc. Hence, to account for these variations and to provide accurate results each individual clinical diagnostic analyzer must be calibrated.

Another problem presented by fluorometers is that the introduction of a new fluorescence label having differing absorption and emission spectra will require a total re-calibration of the entire clinical diagnostic analyzer population. SUMMARY OF THE INVENTION

One object of the present invention is to enable a population of clinical diagnostic analyzers or instruments to be normalized to a specific master clinical diagnostic analyzer or instrument such that the response of any subordinate clinical diagnostic analyzer or instrument in the population to a sample having a specific amount of analyte is substantially the same as the response of the master clinical diagnostic analyzer or instrument to that sample after an initial factory calibration. Another object of the present invention is allow the introduction of a new fluorescence label having a different adsorption and emission spectrum as compared to a prior fluorescence label such that the re-calibration and re-normalization of the entire population of clinical diagnostic analyzers to the master clinical diagnostic analyzer depends only upon the absorption and emission spectra of the new fluores- cence label. Total re-calibration of the population of clinical diagnostic analyzers is not required.

The foregoing and further objects of the invention are accomplished according to one aspect of the invention that provides a method of normalizing a first diagnostic result of a subordinate clinical diagnostic analyzer to a second diagnostic result of a master clinical diagnostic analyzer comprising the steps of obtaining a normalized excitation intensity spectrum of the master clinical diagnostic analyzer, obtaining a normalized excitation intensity spectrum of the subordinate clinical diagnostic analyzer, obtaining a normalized responsivity intensity spectrum of the master clinical diagnostic analyzer, obtaining a normalized responsivity intensity spectrum of the subordinate clinical diagnostic analyzer, obtaining a normalized excitation/emission spectrum of a solid inorganic photostable fluorophore calibration target, reading the solid inorganic photostable fluorophore calibration target in the master clinical diagnostic analyzer thereby obtaining a first response value, reading the solid inorganic photostable fluorophore calibration target in the subordinate clinical diagnostic analyzer thereby obtaining a second response value, determining the gain ratio of the master clinical diagnostic analyzer to the subordinate clini- cal diagnostic analyzer based upon the two above obtained response values, determining a multiplicative normalization factor between a normalized subordinate clinical diagnostic analyzer and the master clinical diagnostic analyzer, determining the relative adsorption/emission spectrum of a first fluorescently labeled dye whereas the first fluorescently labeled dye is a diagnostic assay component, obtaining a first diagnostic result from a specific patient specimen or sample incorporating the first fluorescently labeled dye using the normalized subordinate clinical diagnostic analyzer, and multiplying the first diagnostic result by the multiplicative normalization factor to obtain a second diagnostic result whereas the second di- agnostic result is a normalized approximation to a diagnostic result which would be obtained by analyzing the specific patient specimen or sample on the master clinical diagnostic analyzer.

Still another aspect of the invention provides a method to re-normalize a subordi- nate clinical diagnostic analyzer assay result as compared to a master clinical diagnostic analyzer assay result comprising the steps of normalizing the subordinate clinical diagnostic analyzer as above, obtaining a relative adsorption/intensity spectrum of a second fluorescently labeled dye whereas the second fluorescently labeled dye is a diagnostic assay component, determining a re-normalization mul- tiplicative factor between a subordinate clinical diagnostic analyzer and a master clinical diagnostic analyzer, obtaining a first diagnostic result from a specific patient specimen or sample incorporating the second fluorescently labeled dye using the normalized subordinate clinical diagnostic analyzer, and multiplying the first diagnostic result by the re-normalization factor to obtain a second diagnostic result whereas the second diagnostic result is a normalized approximation to a diagnostic result which would be obtained by analyzing the specific patient specimen or sample on the master clinical diagnostic analyzer.

Further objects, features and advantages of the present invention will be apparent to those skilled in the art from detailed consideration of the preferred embodiments that follow. BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a view of the chemistry associated with an immunoassay usually contained in an analytic slide (not shown).

FIG. 2 is a schematic diagram of an optical detection system associated with a fluorescence detection analyzer.

FIG. 3 is a graph of the relative adsorption and emission spectrum of Alexa Fluor® 635, a dye commonly used a fluorophore label.

FIG. 4 is a graph of the relative adsorption and emission spectrum of Alexa Fluor® 647, another dye commonly used a fluorophore label. FIG. 5 is a graph of an example master instrument linear calibration curve.

FIG. 6 is a graph of an example master instrument nonlinear calibration curve.

FIG. 7 is a histogram of the responses to a set of fixed analyte concentration sam- pies or specimens using Alexa Fluor® 647 when subjected to instrument variations.

FIG. 8 is a histogram of the ratio of responses to a set of fixed analyte concentration samples or specimens comparing bismuth-doped phosphor glass and Alexa Fluor® 647 when subjected to instrument variations.

FIG. 9 is a histogram of the ratio of responses to a set of fixed analyte concentration samples or specimens comparing bismuth-doped phosphor glass and Alexa Fluor® 647 when subjected to instrument variations when corrected using instru- ment normalization. DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

While the present invention is described with respect to preferred embodiments as detailed below and shown in the figures, the present invention is limited only by the metes and bounds of the claims that follow.

Fluorometry is chosen for its extraordinary sensitivity, high specificity, simplicity, and low cost as compared to other analytical techniques. Fluorometry is ordinarily 1000-fold more sensitive than absorbance measurements. It is a widely accepted and powerful technique that is used for a variety of environmental, industrial, and biotechnology applications. It is a valuable analytical tool for both quantitative and qualitative analysis. However, fluorometry requires a stable fluorescence standard to insure that the clinical diagnostic analyzers remain normalized and in calibration. Organic fluorophores, especially those in liquid form, are not well suited for use in normalizing analyzers in a factory setting because they photobleach, have limited shelf life, are prone to carryover problems, and are difficult to dose. A solid inorganic photostable fluorophore would not have the above problems, but there are only a limited number of these materials available. In conjunction with a preferred embodiment of the present inventive method, a device has been designed using a National Institute of Standards and Technology (NIST) developed material consisting of a phosphate matrix glass doped with bismuth ion such that the glass has fluorescent properties. This material is known as NIST Standard Reference Material (SRM) 2944 glass. The composition of such material is shown in Table A.

Table A: Composition of NIST SRM 2944 Glass

This device has been designed to overcome the limitations listed above and is used in connection with this inventive method; see copending United States patent application by Freeman III, Heavner, and Oenick entitled "Fluorescence Reference Standard Device" (Attorney Docket No. CDS5170WOPCT) which is hereby incorporated by reference in its entirety. For different wavelength fluorometry a different material other than NIST SRM 2944 would be used, such as other phosphate doped glasses also available from NIST, including SRM 2943, copper doped glass, spectral correction standard for blue fluorescence.

The above described NIST SRM 2944 glass device is a preferred solid inorganic photostable fluorophore used in the inventive method described herein to normal- ize a population of subordinate clinical diagnostic analyzers to a master clinical diagnostic analyzer. Using the excitation and emission spectrums of the NIST SRM 2944 glass, the excitation and emission spectrums of the fluorescence label employed in the combination fluorescently labeled label antibody reagent, and the measured excitation and responsivity spectrums inherent in the optical detection systems of the master and subordinate clinical diagnostic analyzers, normalization of subordinate analyzers to the master analyzer is performed at the factory. Using samples or specimens of known analyte concentrations, a standard calibration can likewise be performed at the factory. And furthermore, should it be required or desirable to change the label in the combination fluorescence label antibody reagent, this can be accomplished in the field using only the excitation and emission spectrums of the new fluorescence label. One advantage of the inventive method is that by using solid inorganic photostable fluorophore, such as the preferred NIST SRM 2944 glass, as a reference material, a population of subordinate clinical diagnostic analyzers can be normalized to one master clinical diagnostic analyzer such that after a factory normalization and calibration the subordinate clinical diagnostic analyzers will have substantially the same response to a sample or specimen containing a fixed amount of analyte as would the master clinical diagnostic analyzer. Furthermore, should it become necessary or desirable to change the fluorescence label in the combination fluorescence label antibody reagent, then the population of subordinate clinical diagnostic analyzers can be re-normalized (and retain the original factory calibration) by a simple procedure not requiring a total recalibration in the field.

For a general understanding of the disclosed methods, reference is made to the drawings. In the drawings, like reference numerals have been used to designate identical elements. In describing the disclosed methods, the following term(s) have been used in the description.

The term "ξ" (the Greek letter xi) or "emission" refers to one or more wavelengths of light generated as a result of fluorescence, specifically when "ξ" is used in an equation it stands for emission wavelength. The term "responsivity" refers to the normalized output of an optical intensity measuring system as a function of a specific wavelength of light being input to that system.

The term "χ" (the Greek letter chi) or "excitation" refers to one or more wavelengths of light generated to be used as a source to radiate a fluorescence complex, specifically when "χ" is used in an equation it stands for excitation wavelength. The term "absorbance" refers to the normalized extinction coefficient of a fluorescent dye. The term "spectral distribution" or "shape function" refers to the relative intensity of an excitation or emission light beam as a function of wavelength.

The term "clinical diagnostic analyzer," "diagnostic analyzer," and "instrument" are taken to mean devices that accept a patient sample or specimen, analyze the sample or specimen for a specific analyte, and report the result of that analysis. These terms are meant to encompass clinical chemistry analyzers, immunohema- tology analyzers, lateral flow device readers, and the like.

The term "normalize" refers to the inventive method applied to two clinical diagnos- tic analyzers or instruments, a master instrument "A" and a subordinate instrument "B", such that the response of "B" to a specific sample or specimen containing a certain concentration of analyte can be converted to the response of "A" to the same sample or specimen by using a multiplicative factor when the assay method employed by the analyzers uses a common fluorescently labeled dyes.

The term "re-normalize" refers to the inventive method applied to two clinical diagnostic analyzers or instruments, a master instrument "A" and a subordinate instrument "B", such that the response of "B" to a specific sample or specimen containing a certain concentration of analyte can be converted to the response of "A" to the same sample or specimen by using a multiplicative factor when the assay method employed by the analyzers uses differing fluorescently labeled dyes.

The terms "Alexa Fluor® 635" and "Alexa Fluor® 647" refer to preferred organic fluorophores that can be used as fluorescent tags. These materials are made by INVITRO-GEN™. For example, the adsorption/emission spectrum of "Alexa Fluor® 635" is shown in FIG. 3 and "Alexa Fluor® 647" has a absorption maximum at 650 nm and an emission maximum at 671 nm as shown in FIG. 4. "Alexa Fluor® 635" is sometimes abbreviated as "AF 635" and "Alexa Fluor® 647" is sometimes abbreviated as "AF 647" In FIG. 1 a combination fluorescently labeled antibody reagent 101 is added to a target analyte 102 (an antigen in this specific example) in the sample or specimen wherein the combination fluorescence label antibody reagent 101 binds to the analyte forming an antibody-analyte complex 103. Unbound combination fluorescence label antibody reagent 101 is subsequently removed. The bound antibody- analyte complex 103 is then exposed to an excitation light of specific wavelength causing a fluorescence emission proportional to the amount of analyte present to be generated shortly thereafter.

In FIG. 2 the bound antibody-analyte complex 103 is captured in a very thin, well defined volume (normally some form of analysis slide, not shown) and presented at the sample plane 201. Excitation light is generated by the light emitting diode (LED) source 208 then collimated by condenser system lenses 210, filtered by the excitation filer 207, shaped by the excitation aperture 206 and the projection lens 211 , redirected by a dichroic mirror 203, and then passed through an objective lens 202 that acts to converge the excitation light rays down to an area appropriate for the very thin, well defined volume. The excitation system components contained in the dashed rectangle are called the excitation arm 209 of the optical detection system. Any captured and tagged analyte in that volume fluoresces, and a portion of that emission is intercepted by the objective lens 202, passed through the dichroic mirror 203, further passed through at least one band pass filter 204, through a detector lens 212, and finally through a detection aperture 213. The emission light making it through the detection aperture 213 strikes the photodetec- tor 205 and generates an electric current which is amplified into a usable signal. At the photodetector 205, the excitation arm 209 of the optical detection system delivers a photon flux (some number of photons per second) with some spectral distribution (i.e., some mix of wavelengths). This can be described by (χ) = <P · s(x) (1 ) where φ is a scalar (units = photons/second) and S(x) is a unitless shape function where the maximum value of S(x) is unity. The magnitude of φ is determined by the output of the LED source 208, the attenuation properties of the filters 207 and, the attenuation properties of the lenses 210, 211 , and 202, , the reflective properties of the dichroic mirror 203, and the position tolerances of the optical elements. The characteristics of S(x) are determined by the spectral properties of the LED source 208 and the transmission spectrum of the filter 207 and the reflective characteristics of the dichroic mirror 203.

If the NIST SRM 2944 glass is exposed to the photon flux Φ(χ) of equation (1 ), a composite emission curve will be obtained that can be approximated by summation over small increments of Δ χ, i.e., the value of S(x) at a particular χ times the normalized emission curve EC _g i _aS s at that wavelength. That is,

∑x S(x) · ECglass(x, ξ) (2)

The fluorescence photo flux Φ(ξ) emitted by the glass can be written as &θ = Ψ · Ψ ^■ ∑¾c S(¾0 · ECglass(x, ξ) (3) where ψ is a scalar that is characteristic of the output of the NIST SRM 2944 glass. The electrical signal (current) that is generated at the detector at a particular emission wavelength χ can be described by

Ε(ξ) = Φ(ξ) · G · S _R ) (4) where G is a constant and S _R G) is a shape function such that the maximum value of S _R (X) is unity. The magnitude of G is determined by the collection efficiency of the optics 202 and 212, the transmission efficiency of the dichroic mirror 203 and emission filters 204. The characteristics of S _R (£) are determined by the spectral characteristics of the dichroic mirror 203, emission filters 204 and the spectral characteristics of the detector (photodiode) 205. The total electrical signal generated is

E = ∑ _ξ G · S _R © · Φ(ξ) = G ·∑ξ S _R © · [ψ · φ ^■ ∑ _χ S(x) · ECglass( _X , ξ, )] (5) or,

E = G · ψ · φ ^■ ∑ _ξ S _R (0 · [∑ _χ S(x) · ECglassOc, ξ)] (6) Suppose there is a master instrument "A" and a subordinate instrument "B" where Instrument "B" is to be normalized to instrument "A." Using eqn. (6), the ratio of the signals (E _A and E _B ) given by the two instruments in response to being presented with an identical NIST SRM 2944 glass target can be written as

EB ^~ IGB-VJ^B-∑ _X S _R B(X)-¾ S _B >ECglassG )]j or, _{G =} . f∑x Wx>¾ S _B ff)-ECglassff ^, x ⁾ ]) where G _R is called the gain ratio. The responsivities of both instrument A and instrument B ( SRA© and S _RB (¾, respectively) can be measured by presenting a constant intensity variable wavelength light source to each instrument in turn, sweeping the source through the range of wavelengths in the transmission band of the emission filters 204 and the dichroic mirror 203 while monitoring the signal generated by the respective instrument, then normalizing that signal by the maximum value obtained during that sweep. The emission spectra of both instruments, ^S A(X) and S _z (x), are easily measured by a spectrometer.

Consider now the case of a fluorescent label, specifically Alexa Fluor® 647, where in FIG. 3 the solid excitation curve 301 is designated by S _DYE (x). The equivalent digital values for S _DYE (x) are presented in Table 4. Also for Alexa Fluor® 647, in FIG. 3 the dashed emission curve 302 is designated by S _DYE (^) . The equivalent digital values for S _DYE (£) are presented in Table 5.

The photon flux Φ _Ο ΥΕ(0 emitted by the fluorescent label (dye) can be written as Φ _0ΥΕ (ξ) = <p _DYE · φ ^■ [∑ _ξ S ^■ S _DYE © ] · S _DYE (x) (9) where <p _DYE is a scalar that is characteristic of the output of the fluorescence label (dye). Rewriting eqn. (6) in terms of the fluorescence label (dye) gives

E = G ^■ φ _ΌΥΕ · φ · {∑ _ξ Ξ _κ (ξ) ^■ [∑ _χ Ξ(χ) ^■ Ξ _ΟΥΕ (χ) ] ^■ Ξ _ΟΥΕ (χ) } (10) and, or.

Therefore, we can transform the response E _B from analyzer "B" to the response E _A that would be seen by the master analyzer "A" using eqn. (12).

This allows introduction of new fluorescence labels (dyes) to subordinate field instruments and allowing those subordinate field instruments to be re-normalized to a master analyzer "A" by simply providing the absorption and emission spectrum of the new fluorescence label (dye) and using eqn. (12).

In summary, the following is conducted in the factory for each subordinate instrument:

Measure S _z (x) and store this information on the instrument.

Measure S _RZ (¾) and store this information on the instrument.

The Gain Ratio of each instrument is determined by scanning the NIST SRN 2944 calibration slide as a target, and then applying eqn. (8). EXEMPLARY EXAMPLE OF FACTORY NORMALIZATION

In this example, Analyzer AP106 is selected as the master instrument and Analyzer AP1 15 is selected as the subordinate instrument. The goal of this factory nor- malization is to determine the relationship between the two analyzers with respect to their individual responses to the same sample. This means that a response to a particular sample for the subordinate instrument can be converted to the response of the master instrument by multiplying the response of the subordinate instrument by the gain ratio and the remainder of the eqn. (12) to the right of G _R (as derived above and to be determined for this example below). The initial data gathering steps can be listed as follows:

1 . Obtain the normalized excitation intensity spectrum of AP106 (see Table 1 for the digital spectrophotometric data).

2. Obtain the normalized excitation intensity spectrum of AP1 15 (see Table 1 for the digital spectrophotometric data).

3. Obtain the normalized responsivity intensity spectrum of AP106 (see Table 2 for the digital spectrophotometric data).

4. Obtain the normalized responsivity intensity spectrum of AP1 15 (see Table 2 for the digital spectrophotometric data).

5. Obtain the normalized excitation/emission spectrum of NIST SRM 2944 glass (see Table 3A, 3B, and 3C for the digital spectrophotometric data).

6. Read the NIST SRM 2944 glass as a target in AP106 obtaining the response value for E _A of 2181 .705 relative fluorescence units (RFU).

7. Read the NIST SRM 2944 glass as a target in AP1 15 obtaining the response value for E _z of 2035.274 RFU.

8. Calculate the Gain ratio of AP106 to AP1 15 using eqn. (8) where the numerator in brackets has the value 123.9541 and the denominator in brackets has the value 126.4753 with a result as follows: 2181.705 123.9541

= 1.0506

2035.274. 126.4763

Note that the ratio (123.9541/126.4763) expresses the difference in signal between the two instruments based on spectral differences. Whereas the gain ratio G _R expresses differences due to non-spectral differences (e.g., one instrument may have a slightly brighter illumination LED or somewhat more efficient receiver optics).

9. Obtain the relative absorption spectrum of the fluorescent label (dye) used in the analysis. In this case Alexa Fluor® 647 is being used and the associated digital spectrophotometric data is presented in Table 4.

10. Obtain the relative emission intensity of the fluorescent label (dye) used in the analysis. In this case Alexa Fluor® 647 is being used and the associated digital spectrophotometric data is presented in Table 5.

1 1 . Using eqn. (12), the normalization factor between AP1 15 responses and AP106 responses can be determined as follows where the numerator in eqn. (12) is 122.1005 and the denominator in eqn. (12) is 1 17.6860:

That is, to convert a response from AP1 15 to a response normalized to AP106 when making measurements with Alexa Fluor® 647, we must multiply the AP1 15 responses by 1 .0126.

EXEMPLARY EXAMPLE OF FIELD RE-NORMALIZATION

In this example, subordinate instrument AP1 15 has been previously normalized to master instrument AP106 and it is desired to introduce a new fluorescent label (dye). The prior analysis allows introduction of new fluorescent labels (dyes) to subordinate field instruments and allowing those subordinate field instruments to normalize to the master analyzer by simply providing the absorption and emission spectrum of the new dye and using eqn. (12) above. The method is outlined as follows:

1 . Obtain the relative absorption spectrum of the fluorescent label (dye) used in the analysis. In this case Alexa Fluor® 635 is being used and the associated digital spectrophotometric data is presented in Table 6.

2. Obtain the relative emission intensity of the fluorescent label (dye) used in the analysis. In this case Alexa Fluor® 635 is being used and the associated digital spectrophotometric data is presented in Table 7.

3. Using eqn. (12), the normalization factor between AP1 15 responses and AP106 responses can be determined as follows where the numerator in eqn. (12) is 44.02245 and the denominator in eqn. (12) is 45.2193:

That is, to convert a signal from AP1 15 to a signal normalized to AP106 when making measurements with Alexa Fluor® 635 as opposed to Alexa Fluor® 647, we must multiply the AP1 15 signals by 1 .01274. Note that the Gain Ratio is not dependent upon the fluorescent label (dye) and remains constant.

In practice, when a new fluorescence label (dye) is introduced to a subordinate field instrument, the quantity will be provided to that instrument along with

^DYE (0 and ^DYE (X) so that it is capable of running assays that make use of that new fluorescence label (dye). EXEMPLARY EXAMPLE OF FACTORY LINEAR CALIBRATION

In this example, a standard calibration procedure will be conducted with samples or specimens of known analyte concentration. The procedure will utilize 10 samples having known analyte concentrations of 1 , 2, 3, 4, 5, 6, 7, 8, 9, and 10 ng/mL. The master instrument responses were observed to be 1 .15, 1 .90, 3.10, 3.90, 5.05, 5.95, 7.30, 7.90, 8.90, and 10.20. This data is presented in Table 8. For flu- orometry, the amount of emitted light is frequently proportional to amount of analyte present in the sample or specimen; hence, a linear calibration curve is usually employed. Here the known concentration values, which are known without error, are used as the predictor variable and the master instrument responses, containing measurement error, are used as the response variable. This situation is known to satisfy the requirement for using ordinary least squares (OLS) regression. The data of Table 8 and the fitted regression line are shown in Figure 5. The data points 501 of Table 8 and the fitted OLS regression line (shown as a sol- id black line) 502 indicate a close correlation for the master instrument.

For the master instrument, the instrument response to a sample or specimen (as indicated by a y-axis value) is traced back to the right horizontally to the fitted linear calibration line and then traced vertically to the x-axis to obtain the estimate of the analyte concentration in the sample or specimen. For example, in FIG. 5, if the master instrument response to a sample or specimen is 5, then tracing to the right, the horizontal line intercepts the linear calibration line and tracing down a value of approximately 5 on the x-axis is obtained. For the subordinate instruments, subordinate instrument response to a sample or specimen is multiplied by the Gain Ratio and everything to the right of E _z in eqn. (1 ) such that the resulting response can be used just as if it were obtained from the master instrument to obtain an estimate of the analyte concentration in the sample or specimen. For this specific case, the subordinate instrument would produce a response of 4 and subsequently that response would be multiplied by the gain factor of 1 .25 to yield an equivalent master instrument response of 5. Also using the master calibration curve produces an estimated analyte concentration of approximately 5.

EXEMPLARY EXAMPLE OF FIELD LINEAR RE-CALIBRATION

In this example, a new fluorescent label (dye) has been introduced into the analysis chemistry and the subordinate instruments in the field have been re- normalized. For this situation there are two methods that can be used to obtain appropriate estimates of the analyte concentration in the sample or specimen as follows:

1 . The subordinate instrument response can be multiplied by the normalization factor and the old factory calibration curve can be used. In FIG. 5, for a subordinate instrument response of 4 and a normalization factor of 1 .25, this is represented by the up arrow indicating that 4 ^* 1 .25 = 5 is the equivalent master instrument response. Starting a 5 on the y-axis and tracing right to the master linear calibration line 502 and then down to the x-axis produces an estimate of 5.

2. Alternatively, a new linear calibration 503 curve can be constructed by mul- tiplying the slope of the old linear calibration curve 502 by the inverse of the normalization factor. The new calibration curve 503 would then, as in in FIG. 5, have a slope of 1 .005 ^* (1/1 .25) = 0.804. The estimate of the analyte concentration in the sample or specimen would then be obtained by starting on the y-axis at the subordinate instrument response (4) then trac- ing to the right until the new linear calibration curve is encountered and then moving down to the x-axis. It is clear from FIG. 5 that this procedure or the procedure 1 above that both result in the same analyte concentration estimates.

EXEMPLARY EXAMPLE OF FACTORY NON-LINEAR CALIBRATION

In this example, the amount of emitted light is not proportional to amount of analyte present in the sample or specimen; hence, a linear calibration curve cannot be used. In a manner similar to the linear calibration example, a series of 10 samples or specimens of known analyte concentrations is prepared and a non-linear sig- moidal function 601 is fitted to the data points 602. Hence, for the master instrument, a y-axis response is converted into an estimate of analyte concentration, by starting at the y-axis value, tracing right until the calibration curve is encountered and then tracing down to the x-axis to obtain the analyte concentration estimate. EXEMPLARY EXAMPLE OF FIELD NON-LINEAR RE-CALIBRATION

In this example, FIG. 6 shows a non-linear (s-shaped) calibration curve 601. Assuming that the normalization factor is 1 .25 between the master and subordinate instruments, a subordinate instrument response of 4 would generate an equivalent master instrument response of (4 ^* 1 .25) = 5 which would produce an estimated analyte concentration of approximately 5 ½. Unlike the linear calibration case, the slope of the non-linear calibration curve cannot be easily adjusted such that the raw subordinate instrument response can be used. Here it is necessary to multiply the subordinate instrument response by the normalization factor to obtain the equivalent master instrument response and then use the master instrument calibration curve.

SIMULATION TEST OF NORMALIZATION FACTOR EFFECTIVENESS To test the effectiveness of the normalization process, an initial Monte Carlo simulation was conducted where 10,000 simulated fluorescence analyzers were pre- sented with a fixed amount of Alexa Fluor® 647 (AF 647) fluorescence label (dye) or a fixed amount of bismuth-doped glass as used in the NIST SRM 2944 standard. Sources of variation were as follows: 1 . Incubation temperature was allowed to vary between 36° C and 38° C

2. The bandpass characteristics of 3 optical filters in the detection arm of the optical detection system were allowed to vary (according the manufacturers specification)

3. The excitation wavelength was allowed to vary between 630 nm and 636 nm.

After 10,000 simulated analyzers were configured according to the above, the model generated the statistics of a variety of responses. FIG. 7 shows the resulting histogram 701 of responses to simulated samples using AF 647 dye. There is about a +/- 25% range in the population response. FIG. 8 shows the population of ratios of each individual analyzer's response to AF 647 and the NIST SRM 2944 standard. The resulting histogram of responses 801 shows a spread of about +/- 20%. Clearly, the raw response of the NIST SRM 2944 does a poor job in predicting the response of the instrument to the samples using AF 647. However, if the same analysis is repeated where a normalization factor is employed and measurement noise (error) in the following is allowed:

1 . Noise associated with the analyzer aligning and reading the calibration slide. A value was used of 0.5% CV that should be achievable with 4 load and align events each with 4 reads of the fluorescent label (dye). The alignment process was actually a fairly significant source of variability.

2. Noise associated with the characterization of the analyzer's excitation and responsivity spectra by factory calibration instrumentation. It was estimated that each data point had a 0.25% CV imprecision.

3. Noise associated with the characterization of the analyzer's incubation temperature. Noise of one standard deviation was taken to be 0.067° C. This creates an error because of the temperature sensitivity differences between the NIST SRM 2944 glass (—0.25% per °C) and the AF 647 fluorescent label (dye) (-1 .2% per °C). FIG. 9 contains the resulting histogram of errors 901 which indicates that the overall analyte estimation error has been reduced to about a range of +/- 1 %.

It will be apparent to those skilled in the art that various modifications and variations can be made to the article of manufacture disclosed herein. Thus, it is in- tended that the present invention cover such modifications and variations, provided they come within the scope of the appended claims and their equivalents.

The disclosure of all publications cited above is expressly incorporated herein by reference in their entireties to the same extent as if each were incorporated by reference individually.

Table 1 - Nornnalized Excitation Intensity of Analyzers AP106 and AP1 15

Table 2 - Nornnalized Responsivity of Analyzers AP106 and AP1 15

Table 3A - Normalized Excitation/Emission Spectrum of NIST SRM 2944 Glass

(Excitation Range of 630 nm to 638 nm)

Table 3B - Normalized Excitation/Emission Spectrum of NIST SRM 2944 Glass

(Excitation Range of 639 nm to 647 nm)

Table 3C - Normalized Excitation/Emission Spectrum of NIST SRM 2944 Glass

(Excitation Range of 648 nm to 655 nm)

Table 4 - Relative Absorption of Alexa Fluor® 647 vs. Excitation Wavelength

Table 5 - Relative Emission Intensity of Alexa Fluor® 647 vs. Wavelength

701 0.3303

702 0.3180

703 0.3064

704 0.2973

705 0.2845

706 0.2749

707 0.2693

708 0.2625

709 0.2528

710 0.2494

Table 6 - Relative Absorption of Alexa Fluor® 635 vs. Excitation Wavelength

337 0.037430235

338 0.037647515

339 0.038295235

340 0.038718467

341 0.040220889

342 0.041395847

343 0.042836484

344 0.043783865

345 0.045607573

346 0.04683196

347 0.048283924

348 0.049325013

349 0.050344477

350 0.051013822

351 0.052280429

352 0.053093941

353 0.054082512

354 0.055019595

355 0.056027732

356 0.057377749

357 0.058983148

358 0.060467034

359 0.061568879

360 0.063565587

361 0.06554273

362 0.067015289

363 0.068705128

364 0.070186955

365 0.071350586

366 0.072462729

367 0.073389514

368 0.073986776

369 0.074636556

370 0.07504846

371 0.075469633

372 0.07612868

373 0.076315067

374 0.075985544

375 0.076880406

376 0.076726972

377 0.07677846

378 0.077796894 379 0.078806061

380 0.079362132

381 0.081236299

382 0.082286655

383 0.083790108

384 0.085303857

385 0.086354214

386 0.086777446

387 0.086087506

388 0.085169988

389 0.083501774

390 0.081287787

391 0.078620704

392 0.075923758

393 0.072740764

394 0.069394039

395 0.065687926

396 0.062051838

397 0.058406481

398 0.055009297

399 0.05188809

400 0.048953269

401 0.046184239

402 0.044638567

403 0.042949758

404 0.042712913

405 0.042549181

406 0.043176305

407 0.044144281

408 0.045503567

409 0.052867393

410 0.048532096

411 0.049653507

412 0.051084875

413 0.051868524

414 0.052475054

415 0.052331917

416 0.051981798

417 0.051003524

418 0.049582453

419 0.047851424

420 0.04626662 421 0.044124716

422 0.041313467

423 0.039376485

424 0.036791783

425 0.034980432

426 0.032899284

427 0.031241367

428 0.029399123

429 0.027843153

430 0.027060535

431 0.026195535

432 0.024899065

433 0.024362559

434 0.023754999

435 0.023806488

436 0.022746863

437 0.022489422

438 0.021737696

439 0.021623393

440 0.02098494

441 0.020500952

442 0.02033722

443 0.020048887

444 0.019687441

445 0.019410435

446 0.018935715

447 0.018514542

448 0.018081012

449 0.017495078

450 0.01718512

451 0.017031685

452 0.016351013

453 0.015918513

454 0.015301686

455 0.015095733

456 0.014672501

457 0.014488174

458 0.014127757

459 0.013766311

460 0.013405894

461 0.013417222

462 0.012995019 463 0.012892043

464 0.012686091

465 0.012531627

466 0.012098097

467 0.011984823

468 0.011593514

469 0.011378293

470 0.01108996

471 0.010760436

472 0.010460776

473 0.010513294

474 0.010307341

475 0.010163175

476 0.010060198

477 0.010019008

478 0.009782163

479 0.009524722

480 0.009410419

481 0.009380556

482 0.00910252

483 0.008936728

484 0.008947026

485 0.00884405

486 0.008721508

487 0.008752401

488 0.008833752

489 0.008927461

490 0.009071627

491 0.009349663

492 0.009452639

493 0.009668889

494 0.009957222

495 0.010213633

496 0.010285716

497 0.010523591

498 0.010862383

499 0.011192936

500 0.01126502

501 0.011512162

502 0.011995121

503 0.012314347

504 0.012850853 505 0.013324543

506 0.013818829

507 0.014385198

508 0.014941269

509 0.015310953

510 0.016062679

511 0.016670239

512 0.01765881

513 0.01830859

514 0.019122102

515 0.020306327

516 0.021489524

517 0.022808648

518 0.024063928

519 0.025599303

520 0.027102755

521 0.028935731

522 0.030593647

523 0.032591385

524 0.034639581

525 0.036410771

526 0.038831741

527 0.040675014

528 0.042887972

529 0.045257454

530 0.047501305

531 0.049591721

532 0.05193031

533 0.054257571

534 0.056286202

535 0.05850019

536 0.060364058

537 0.062247492

538 0.06398779

539 0.065666301

540 0.066943206

541 0.068456956

542 0.069580426

543 0.070702866

544 0.072215586

545 0.073399812

546 0.074913562 547 0.076458204

548 0.078312805

549 0.079991317

550 0.082370066

551 0.085067012

552 0.087806178

553 0.091400047

554 0.095117486

555 0.099360105

556 0.104055818

557 0.109421907

558 0.115342007

559 0.121850101

560 0.128842183

561 0.136432556

562 0.144618132

563 0.153289756

564 0.162844915

565 0.172514377

566 0.18313225

567 0.193594629

568 0.204994091

569 0.216249387

570 0.228512819

571 0.240241805

572 0.25253716

573 0.264133307

574 0.275821102

575 0.287219535

576 0.298712705

577 0.309184352

578 0.319399588

579 0.328667444

580 0.3370703

581 0.344207578

582 0.350303767

583 0.354978886

584 0.358552159

585 0.359890849

586 0.360209045

587 0.358386367

588 0.355236326 589 0.350581803

590 0.345050953

591 0.338203037

592 0.330666211

593 0.322880183

594 0.315260976

595 0.307969233

596 0.301574413

597 0.296158896

598 0.292213879

599 0.289752749

600 0.289135921

601 0.289979296

602 0.292976933

603 0.297548045

604 0.304026276

605 0.312582566

606 0.323190141

607 0.336041568

608 0.350643589

609 0.367674818

610 0.386478267

611 0.407578084

612 0.430449092

613 0.455286945

614 0.48257666

615 0.510771536

616 0.541385322

617 0.573123608

618 0.607094417

619 0.641138339

620 0.677139839

621 0.7121816

622 0.7481831

623 0.78335873

624 0.818423146

625 0.852023241

626 0.883297105

627 0.911862695

628 0.937339

629 0.959623043

630 0.977274189 631 0.989372859

632 0.998094941

633 1

634 0.996438054

635 0.98686127

636 0.97186794

637 0.951200622

638 0.924982889

639 0.893842894

640 0.858419091

641 0.820461045

642 0.778570338

643 0.734971257

644 0.6888472

645 0.643063994

646 0.595664062

647 0.549468951

648 0.504469394

649 0.461135991

650 0.419945523

651 0.379960905

652 0.343331251

653 0.308186513

654 0.276006459

655 0.245730435

656 0.218813494

657 0.193428838

658 0.171083008

659 0.150746244

660 0.132704819

661 0.116547858

662 0.102007622

663 0.089783321

664 0.078539353

665 0.068292194

666 0.059540249

667 0.052105369

668 0.04514418

669 0.039252913

670 0.034093807

671 0.029614344

672 0.025949422 673 0.02249869

674 0.019667875

675 0.017217042

676 0.015260495

677 0.01325246

678 0.011459645

679 0.01014155

680 0.009071627

681 0.007897699

682 0.006847342

683 0.005826848

684 0.005076152

685 0.004405777

686 0.003849705

687 0.00357167

688 0.002840539

689 0.002572801

690 0.002161926

691 0.001851968

692 0.001502879

693 0.001399902

694 0.00119292

695 0.000894289

696 0.000606986

697 0.00050298

698 0.000420599

699 0.000400004

700 6.01824E-05

701 0

Table 7 - Relative Emission of Alexa Fluor® 635 vs. Excitation Wavelength

641 0.901883

642 0.932834

643 0.954136

644 0.979742

645 0.989957

646 0.999566

647 0.999695

648 1

649 0.996131

650 0.978916

651 0.970612

652 0.950181

653 0.925314

654 0.897187

655 0.8698

656 0.834891

657 0.802113

658 0.766769

659 0.734991

660 0.704212

661 0.676042

662 0.635264

663 0.602313

664 0.5681

665 0.541495

666 0.51002

667 0.480893

668 0.454419

669 0.428379

670 0.402556

671 0.378646

672 0.357867

673 0.34013

674 0.318828

675 0.302047

676 0.285789

677 0.26753

678 0.251533

679 0.242272

680 0.22684

681 0.218666

682 0.208494 683 0.195713

684 0.187975

685 0.181454

686 0.172065

687 0.163109

688 0.157284

689 0.150806

690 0.143981

691 0.139417

692 0.134678

693 0.131461

694 0.127027

695 0.123636

696 0.121288

697 0.119202

698 0.117637

699 0.114681

700 0.114072

701 0.110768

702 0.107812

703 0.107551

704 0.106508

705 0.105378

706 0.103986

707 0.101465

708 0.099683

709 0.099074

710 0.098596

711 0.097987

712 0.09577

713 0.094422

714 0.094118

715 0.090597

716 0.087076

717 0.083772

718 0.085076

719 0.082381

720 0.081294

721 0.078207

722 0.075294

723 0.074208

724 0.070556 725 0.069165

726 0.067513

111 0.063253

728 0.058731

729 0.058949

730 0.056384

731 0.055297

732 0.052341

733 0.049385

734 0.046211

735 0.045081

736 0.044864

737 0.040082

738 0.039864

739 0.037778

740 0.03643

741 0.034778

742 0.032952

743 0.029996

744 0.029474

745 0.028605

746 0.025736

747 0.025301

748 0.023649

749 0.023867

750 0.021475

751 0.022388

752 0.019389

753 0.019389

754 0.018215

755 0.016302

756 0.016694

757 0.016346

758 0.015694

759 0.015215

760 0.014737

761 0.012911

762 0.012129

763 0.012042

764 0.012737

765 0.011172

766 0.010999 767 0.009868

768 0.009912

769 0.010607

770 0.010477

771 0.008695

772 0.008868

773 0.008868

774 0.008129

775 0.008955

776 0