# US 2016/0259792 A1 · Optimized spectral matching and display

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2016/0259792 A1
- Document ID: US20160259792A1
- Status: Published application
- Kind: A1
- Application number: 15/060,260
- Filed: 2016-03-03
- Published: 2016-09-08
- Earliest claimed: 2015-03-05
- Assignees: Bio-Rad Laboratories, Inc.
- Applicants: Bio-Rad Laboratories, Inc.
- Inventors: Karl NEDWED; Ty ABSHEAR
- Web page: https://patentgrep.com/patent/US20160259792A1

## CPC

- G01N 21/35: using infrared light (G01N21/39 takes precedence)
- G06F 16/24578: {using ranking}
- G06F 16/248: Presentation of query results
- G06F 18/22: Matching criteria, e.g. proximity measures
- G06F 18/23: Clustering techniques
- G06T 11/23: using straight lines or curves
- G06V 10/60: relating to illumination properties, e.g. using a reflectance or lighting model
- G01N 21/552: Attenuated total reflection
- G01N 2201/121: Correction signals
- G01N 2201/127: Calibration; base line adjustment; drift compensation
- G01N 2201/129: Using chemometrical methods
- G06V 2201/03: Recognition of patterns in medical or anatomical images
- G01N: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES (measuring or testing apparatus or processes other than immunoassay, involving enzymes or microorganisms C12M, C12Q)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

Systems, methods, and apparatuses are provided for identifying an optimal spectral match and potentially display the compared spectra. A sample spectrum of a sample substance can be compared to reference spectra to identify matches, thereby determining possibilities for what the sample substance is. Correction parameter(s) may be used for the sample spectrum and/or the reference spectrum. Initial value(s) for the correction parameter(s) can be applied to the sample spectrum and/or a reference spectrum, and a similarity score can be determined. The value(s) for the correction parameter(s) can be updated and iteratively improved to provide an optimal similarity score that satisfies a convergence criterion. Data about the reference substances having optimal similarity scores that are above a threshold can be output to a user, e.g., the reference spectra can overlay the sample spectrum. A user can then make a final determination of which reference substance corresponds to the sample substance.

## Claims

### Claim 1

A method for identifying a reference substance that matches a sample substance by comparing spectra, the method comprising performing, by a computer system:
receiving a sample spectrum of the sample substance, the sample spectrum having an intensity value for each of a plurality of discriminating values of a discriminating property;
for each of a plurality of reference substances:
retrieving, from a database, a reference spectrum for the respective reference substance, the reference spectrum having intensity values for the plurality of discriminating values of the discriminating property;
initially selecting one or more correction values for one or more correction parameters;
for each of a plurality of iterations:
applying the one or more correction values for the one or more correction parameters to at least one of the sample spectrum and the reference spectrum;
computing a similarity score between the sample spectrum and the reference spectrum resulting from application of the one or more correction values;
determining whether the similarity score satisfies one or more convergence criteria;
when the similarity score satisfies the one or more convergence criteria, identifying the similarity score as an optimized similarity score corresponding to one or more optimized values of the one or more correction parameters;
when the similarity score does not satisfy the one or more convergence criteria, updating the one or more correction values for use in performing another iteration;
comparing the optimized similarity score to a threshold to determine whether the reference substance is a potential match for the sample substance; and
outputting data about one or more of the plurality of reference substances that have optimized similarity scores that are above the threshold.

### Claim 2 (depends on 1)

The method of claim 1, wherein the one or more correction parameters are a plurality of correction parameters, thereby providing a set of corrections values, and wherein updating the one or more correction values for use in performing the other iteration includes:
for each correction parameter of the plurality of correction parameters:
determining an optimal value for the correction parameter while keeping other correction values fixed, wherein the optimal value is based on optimizing the similarity score; and
updating the plurality of correction values to have the optimal values.

### Claim 3 (depends on 2)

The method of claim 2, wherein the optimal values of the one or more correction parameters are determined sequentially, and wherein the optimal value of a next correction parameter is determined while keeping any previous correction parameters constant at a previously determined optimal value.

### Claim 4 (depends on 2)

The method of claim 2, wherein the plurality of corrections parameters includes a first correction parameter, and wherein determining the optimal value for the first correction parameter includes:
for each of a plurality of subiterations:
computing the similarity score between the sample spectrum and the reference spectrum resulting from application of the set of correction values at each of a plurality of correction values of the first correction parameter;
identifying a top similarity score at the sets of correction values;
determining whether the top similarity score satisfies one or more other convergence criteria;
when the top similarity score satisfies the one or more other convergence criteria, identifying the value for the first correction parameter as the optimal value for the first correction parameter;
when the top similarity score does not satisfy the one or more other convergence criteria, determining new correction values for a next subiteration based on the set of correction values corresponding to the top similarity score.

### Claim 5 (depends on 1)

The method of claim 1, wherein the one or more correction parameters include a clipping parameter that specifies a maximum intensity value at any discriminating value of one of the sample spectrum or the reference spectrum to be used when computing the similarity score.

### Claim 6 (depends on 1)

The method of claim 1, wherein the one or more correction parameters include at least one of: a horizontal shift correction and a vertical offset correction.

### Claim 7 (depends on 1)

The method of claim 1, wherein the one or more correction parameters include a plurality of conversion correction parameters for converting a transmission IR spectrum to an attenuated total reflectance IR spectrum or converting an attenuated total reflectance IR spectrum to a transmission IR spectrum.

### Claim 8 (depends on 7)

The method of claim 7, wherein the plurality of conversion correction parameters include a polarization adjustment parameter P that reduces intensity values above a specified amount for a cluster of discriminating values, wherein the intensity values of the cluster of discriminating values are reduced by a factor that is proportional to a maximum intensity of the cluster of discriminating values.

### Claim 9 (depends on 8)

The method of claim 8, wherein the polarization adjustment parameter P is used to performed a correction of an intensity I at a particular discriminating value using IC=0.5+(I−0.5)·(1−(1−P)·hc), where Ic is a corrected intensity, and hc is a maximum intensity value within a cluster that includes the particular discriminating value.

### Claim 10 (depends on 7)

The method of claim 7, wherein the plurality of conversion correction parameters include a horizontal shift parameter H that is applied to a discriminating value v using vcorr=v+H·hc, wherein vcorr is a corrected discriminating value, and wherein hc is a maximum intensity value within a cluster that includes the discriminating value.

### Claim 11 (depends on 7)

The method of claim 7, wherein the plurality of conversion correction parameters include a peak intensity shift F that is applied an intensity I at a particular discriminating value v using equation IC=I·(1+F·(v−v0)/v0), wherein IC is a corrected intensity, and v0 is a first discriminating value of a first data point in a spectrum to which the plurality of conversion correction parameters are applied.

### Claim 12 (depends on 1)

The method of claim 1, further comprising:
for a first reference spectrum of a first reference substance of the plurality of reference substances:
performing a baseline correction to at least one of the sample spectrum and the first reference spectrum, prior to the plurality of iterations.

### Claim 13 (depends on 12)

The method of claim 12, further comprising:
determining four respective optimized similarity scores for a first reference spectrum corresponding to four options of applying the baseline correction to only the sample spectrum, to only the first reference spectrum, to both the sample spectrum and the first reference spectrum, and not applying the baseline correction to either the sample spectrum or the first reference spectrum; and
using a highest of the four respective optimized similarity scores as the optimized similarity score for the first reference spectrum.

### Claim 14 (depends on 1)

The method of claim 1, wherein the threshold is one of: an absolute number and a rank of the similarity scores.

### Claim 15 (depends on 1)

The method of claim 1, wherein outputting the data about the one or more of the reference substances includes:
plotting curves of the sample spectrum and a first reference spectrum in an overlapping, stacked, or offset fashion using the one or more optimized values of the one or more correction parameters.

### Claim 16 (depends on 1)

The method of claim 1, wherein the discriminating property is wavelength.

### Claim 17

A computer product comprising a non-transitory computer readable medium storing a plurality of instructions, that when executed on one or more processors of a computer system, perform:
receiving a sample spectrum of the sample substance, the sample spectrum having an intensity value for each of a plurality of discriminating values of a discriminating property;
for each of a plurality of reference substances:
retrieving, from a database, a reference spectrum for the respective reference substance, the reference spectrum having intensity values for the plurality of discriminating values of the discriminating property;
initially selecting one or more correction values for one or more correction parameters;
for each of a plurality of iterations:
applying the one or more correction values for the one or more correction parameters to at least one of the sample spectrum and the reference spectrum;
computing a similarity score between the sample spectrum and the reference spectrum resulting from application of the one or more correction values;
determining whether the similarity score satisfies one or more convergence criteria;
when the similarity score satisfies the one or more convergence criteria, identifying the similarity score as an optimized similarity score corresponding to one or more optimized values of the one or more correction parameters;
when the similarity score does not satisfy the one or more convergence criteria, updating the one or more correction values for use in performing another iteration;
comparing the optimized similarity score to a threshold to determine whether the reference substance is a potential match for the sample substance; and
outputting data about one or more of the plurality of reference substances that have optimized similarity scores that are above the threshold.

### Claim 18 (depends on 17)

The computer product of claim 17, wherein the one or more correction parameters are a plurality of correction parameters, thereby providing a set of corrections values, and wherein updating the one or more correction values for use in performing the other iteration includes:
for each correction parameter of the plurality of correction parameters:
determining an optimal value for the correction parameter while keeping other correction values fixed, wherein the optimal value is based on optimizing the similarity score; and
updating the plurality of correction values to have the optimal values.

### Claim 19 (depends on 18)

The computer product of claim 18, wherein the plurality of corrections parameters includes a first correction parameter, and wherein determining the optimal value for the first correction parameter includes:
for each of a plurality of subiterations:
computing the similarity score between the sample spectrum and the reference spectrum resulting from application of the set of correction values at each of a plurality of correction values of the first correction parameter;
identifying a top similarity score at the sets of correction values;
determining whether the top similarity score satisfies one or more other convergence criteria;
when the top similarity score satisfies the one or more other convergence criteria, identifying the value for the first correction parameter as the optimal value for the first correction parameter;
when the top similarity score does not satisfy the one or more other convergence criteria, determining new correction values for a next subiteration based on the set of correction values corresponding to the top similarity score.

### Claim 20 (depends on 17)

The computer product of claim 17, wherein the one or more correction parameters include a clipping parameter that specifies a maximum intensity value at any discriminating value of one of the sample spectrum or the reference spectrum to be used when computing the similarity score.