# US 2021/0239592 A1 · Methods and systems for classifying fluorescent flow cytometer data > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2021/0239592 A1 - Document ID: US20210239592A1 - Status: Published application - Kind: A1 - Application number: 17/153,500 - Filed: 2021-01-20 - Published: 2021-08-05 - Earliest claimed: 2020-01-31 - Assignees: Becton, Dickinson and Company - Applicants: Becton, Dickinson and Company - Inventors: Richard Lee Halpert - Web page: https://patentgrep.com/patent/US20210239592A1 ## CPC - G01N 15/1459: {the analysis being performed on a sample stream} - G01N 15/1429: Signal processing - G01N 15/1434: Optical arrangements - G01N 15/147: {the analysis being performed on a sample stream} - G06F 18/23: Clustering techniques - G06V 10/762: using clustering, e.g. of similar faces in social networks - G06V 10/7715: {Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods} - G06V 20/698: {Matching; Classification} - G01N 2015/1006: {for cytology} - G01N 2015/1402: {Data analysis by thresholding or gating operations performed on the acquired signals or stored data} - G01N 2015/145: {by pattern of light, e.g. fringe pattern} - 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) - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract Methods for classifying fluorescent flow cytometer data are provided. In some instances, methods include processing the flow cytometer data with a supervised algorithm configured to cluster the fluorescent flow cytometer data into distinct populations according to the relationship of data points to relevant threshold values. In embodiments, methods include determining a measure of spillover spreading by calculating spillover spreading coefficients and combining them in a spillover spreading matrix. In some embodiments, populations of fluorescent flow cytometer data are adjusted to subtract the magnitude of spillover spreading. In embodiments, spillover spreading adjusted populations are partitioned after potential partitions are evaluated relative to the threshold values. In embodiments, partitioned populations of fluorescent flow cytometer data are classified (i.e., phenotyped) according to a hierarchy. Systems and computer-readable media for classifying fluorescent flow cytometer data are also provided. ## Claims ### Claim 1 A method of classifying fluorescent flow cytometer data, the method comprising: processing the flow cytometer data with a supervised algorithm configured to: cluster the fluorescent flow cytometer data into populations; determine a measure of spillover spreading for the populations of fluorescent flow cytometer data; adjust the populations of fluorescent flow cytometer data based on the determined spillover spreading to generate distinct spillover spreading adjusted populations; and establish partitions between the distinct spillover spreading adjusted populations of fluorescent flow cytometer data to classify the distinct spillover spreading adjusted populations of fluorescent flow cytometer data. ### Claim 2 (depends on 1) The method according to claim 1, wherein the fluorescent flow cytometer data is clustered into populations based on positivity or negativity of the fluorescent flow cytometer data with respect to particular fluorochromes. ### Claim 3 (depends on 1) The method according to claim 1, wherein the fluorescent flow cytometer data is determined to be positive or negative for a particular fluorochrome based on a relationship of the fluorescent flow cytometer data to a threshold value. ### Claim 4 (depends on 1) The method according to claim 1, wherein determining spillover spreading comprises quantifying the extent to which fluorescent flow cytometer data collected by a fluorescent light detector is increased by the collection of light emitting from a particular fluorochrome. ### Claim 5 (depends on 1) The method according to claim 1, wherein determining spillover spreading comprises calculating a spillover spreading coefficient for a fluorescent light detector-fluorochrome pair. ### Claim 6 (depends on 5) The method according to claim 5, wherein the spillover spreading coefficient is calculated according to Equation 1: SS=(Δσ_ƒ)√Δd=√((σ_pos){circumflex over ( )}2-(σ_neg){circumflex over ( )}2)/√(d_pos-d_neg) wherein: SS is the spillover spreading coefficient; Δσ_f is an incremental standard deviation indicating the spread of the emission between the positive and negative fluorescent flow cytometer data collected from a fluorochrome; and Δd is a difference in the intensity of the fluorescent light between the positive and negative fluorescent flow cytometer data received by a fluorescent light detector. ### Claim 7 (depends on 5) The method according to claim 5, wherein calculating the spillover spreading coefficient comprises assuming that the intensity of fluorescent light collected by the fluorescent light detector for the negative population of flow cytometer data is zero. ### Claim 8 (depends on 7) The method according to claim 7, wherein the spillover spreading coefficient is calculated according to Equation 2: SS_=√(σ{circumflex over ( )}2−σ_0{circumflex over ( )}2)√d wherein: SS is the spillover spreading coefficient; σ{circumflex over ( )}2 is the standard deviation of the positive population of fluorescent flow cytometer data; σ_0{circumflex over ( )}2 is an estimate of the standard deviation of the negative population of fluorescent flow cytometer data; and d is the intensity of light collected by a fluorescent light detector. ### Claim 9 (depends on 8) The method according to claim 8, wherein σ_0{circumflex over ( )}2 is calculated by linear regression. ### Claim 10 (depends on 1) The method according to claim 1, wherein the fluorescent flow cytometer data is collected from light emitting from a plurality of different fluorochromes. ### Claim 11 (depends on 10) The method according to claim 10, wherein the plurality of different fluorochromes ranges from 2 to 20 different fluorochromes. ### Claim 12 (canceled) ### Claim 13 (depends on 10) The method according to claim 10, wherein a spillover spreading coefficient is calculated for each fluorescent light detector-fluorochrome pair. ### Claim 14 (depends on 10) The method according to claim 10, wherein the spillover spreading coefficients calculated for each fluorescent light detector-fluorochrome pair are combined in a spillover spreading matrix. ### Claim 15 (depends on 14) The method according to claim 14, wherein the method further comprises computing the magnitude of spillover spreading for each of the plurality of different fluorochromes based on the spillover spreading matrix. ### Claim 16 (depends on 15) The method according to claim 15, wherein adjusting fluorescent flow cytometer data comprises subtracting the magnitude of the spillover spreading for each of the plurality of different fluorochromes from the populations of fluorescent flow cytometer data corresponding to that fluorochrome. ### Claim 17 (depends on 1) The method according to claim 1, wherein establishing partitions between distinct spillover spreading adjusted populations of fluorescent flow cytometer data comprises evaluating the separation of the distinct spillover spreading adjusted populations relative to a threshold value. ### Claim 18 (depends on 17) The method according to claim 17, wherein evaluating the separation of distinct spillover spreading adjusted populations relative to a threshold value comprises calculating Matthew's correlation coefficient. ### Claim 19 (depends on 18) The method according to claim 18, wherein the fluorescent flow cytometer data does not contain positive data for at least one fluorochrome. ### Claim 20 (depends on 19) The method according to claim 19, wherein evaluating the separation of distinct spillover spreading adjusted populations relative to a threshold value comprises calculating balanced accuracy. ### Claim 21 (depends on 1) The method according to claim 1, wherein the distinct spillover spreading adjusted populations of fluorescent flow cytometer data are partitioned according to a hierarchy. ### Claim at position 22 22-66. (canceled)