Cell analysis method, training method for deep learning algorithm, cell analyzer, training apparatus for deep learning algorithm, cell analysis program, and training program for deep learning algorithm

A blood cell analyzer configured to perform a cell-type-identification of cells contained in a blood sample by using an artificial intelligence algorithm having a neural network structure, the blood cell analyzer comprising:

• a cell measuring unit including:
• a sample preparator configured to prepare a first measurement sample including first cells from the blood sample and to prepare a second measurement sample including second cells from the blood sample, wherein the second measurement sample is prepared by mixing the blood sample, a staining reagent and a hemolytic reagent;
• and
• an electric detector configured to electrically measure the first cells;
• an optical detector comprising (i) a flow cell through which the second cells flow, (ii) a light source configured to irradiate a beam for optical interrogation of the second cells, wherein the beam is directed to the flow cell and
• wherein a set of lights is generated upon interaction of the beam with a respective one of the second cells optically interrogated at the flow cell
• and (iii) a plurality of light detectors configured to respectively sense the set of lights from the respective one of the optically interrogated second cells, wherein the set of lights sensed by the light detectors carries analog waveform signals indicative of different perspectives of a morphological feature of the respective one of the optically interrogated second cells, and wherein the analog waveform signals are converted into a matrix of values digitally representing intensities of the set of lights at multiple time points;
• and
• a processor programmed to execute a cell-type-identification process on the respective one of the optically interrogated second cells by using the artificial intelligence algorithm, wherein the cell-type-identification process comprises:
• receiving the matrix of values;
• and
• performing a calculation on the received matrix of values by using the artificial intelligence algorithm trained in advance to the cell-type-identification process,
• wherein the processor is further programmed to provide an analysis result including (i) red blood cell counts, (ii) platelet counts, and (iii) a white blood cell classification result, wherein the red blood cell counts and platelet counts are derived from an electric measurement of the first cells obtained by the electric detector, and wherein the white blood cell classification result is derived from execution of the cell-type-identification process, the white blood cell classification result including cell counts of subpopulations of white blood cells, and the subpopulations including lymphocytes, monocytes, eosinophils, neutrophils, and basophils.