A bone marrow cell recognition result merging and counting method and system based on a double space ontology constraint model

Abstract

The application discloses a bone marrow cell recognition result merging and counting method and system based on a double-space ontology constraint model. The method acquires a single cell record set, loads a statistical rule model containing a fine-grained label space, a clinical statistical space, a merging relationship, a denominator caliber relationship, an abnormal isolation relationship and a rule version identifier, forms a conversion path constrained by the rule version identifier, after rule integrity verification, maps an original fine-grained category to a clinical statistical node, determines a denominator caliber type and divides a denominator set, and generates an independent abnormal statistical item record; generates a clinical statistical result data set according to the clinical statistical node, the denominator set and the independent abnormal statistical item, and forms a reversible traceable data set. The application can reduce statistical denominator mixing, retain abnormal morphological characteristics, support rule version traceability, model upgrade compatibility and artificial correction incremental recalculation.

Images