US 12,223,753 B2
Cytometry data analysis
Ria Baumgrass, Nuthetal (DE); and Yen Hoang, Berlin (DE)
Assigned to DEUTSCHES RHEUMA-FORSCHUNGSZENTRUM BERLIN, Berlin (DE)
Appl. No. 17/600,131
Filed by DEUTSCHES RHEUMA-FORSCHUNGSZENTRUM BERLIN, Berlin (DE)
PCT Filed Apr. 2, 2020, PCT No. PCT/EP2020/059435
§ 371(c)(1), (2) Date Sep. 30, 2021,
PCT Pub. No. WO2020/201443, PCT Pub. Date Oct. 8, 2020.
Claims priority of application No. 19166801 (EP), filed on Apr. 2, 2019.
Prior Publication US 2022/0207895 A1, Jun. 30, 2022
Int. Cl. G06V 20/69 (2022.01); G01N 15/14 (2024.01); G06N 20/20 (2019.01); G06V 10/82 (2022.01); G06V 10/84 (2022.01)
CPC G06V 20/698 (2022.01) [G01N 15/14 (2013.01); G06N 20/20 (2019.01); G06V 10/82 (2022.01); G06V 10/84 (2022.01)] 13 Claims
OG exemplary drawing
 
1. Method for classifying selected marker signals from cytometric measurements comprising a first measurement and a second measurement, wherein the first measurement comprises a cytometric measurement acquired from a first sample of N1 particles, and the second measurement comprises a cytometric measurement acquired from a second sample of N2 particles, N1 being the number of particles in the first sample and N2 being the number of particles in the second sample, wherein each particle niN1 of the N1 particles of the first sample is labelled with a number of L1 fluorescent, mass or oligo markers liN1, each particle niN2 of the N2 particles of second sample is labelled with a number of L2 fluorescent, r mass or oligo markers liN2, wherein the first measurement acquires for each particle niN1 a detected intensity pljN2 of each marker ljN1 of the particle and the second measurement acquires for each particle niN2 a detected intensity pljN2 of each marker ljN2 of the particle, the method comprising the steps of:
a) binning the detected intensities of each marker ljN1 of the first measurement into an associated first number B11 of bins,
b) generating at least one associated marker function, in particular a ratio of the binned intensities, of two markers or a number of particles per bin of a marker,
c) generating from the first measurement a number of Si first feature sets, each first feature set relating a different combination of two binned marker intensities and the at least one associated marker function or a third marker intensity to each other,
d) binning the detected intensities of each marker ljN2 of the second measurement into an associated second number B12 of bins,
e) generating at least one associated marker function, in particular a ratio of the binned intensities, of two markers or a number of particles per bin of a marker,
f) generating from the second measurement a number of S2 second feature sets, each second feature set relating a different combination of two binned marker intensities and the at least one associated marker function or a third marker intensity to each other, wherein each first feature set and each second feature set comprising the same combination of markers intensities or functions, form a pair of feature sets, and
g) providing the pairs of feature sets to a machine learning method, wherein the machine learning method determines at least one selected pair of feature sets that exhibits the largest variation between the first and the second feature set.