Adherent red blood cell automatic counting method based on high spectral imaging

Abstract

The invention discloses an adherent red blood cell automatic counting method based on high spectral imaging. The method includes the following steps: reading high spectral image data of a blood film and compressing the data, decomposing the compressed high spectral data through a continuous maximum angle convex cone method, and obtaining abundance maps of a preset number of end members of the blood film; conducting binary processing on the abundance maps of the end members through the Otsu method, and removing fine noise spots through erosion operation by a mathematic morphological method; conducting hole filling on binary images of the end members, and selecting a cytoplasm binary image on the basis of the number of connected domains and the size of the maximum connected domain; marking and counting areas of cytoplasm connected domains of the cytoplasm binary image, and selecting a median of the areas of the cytoplasm connected domains as a reference value R; and identifying and counting all the cytoplasm connected domains. Through full utilization of spectrum and image information, the counting problem of adherent red blood cells is solved, and the accuracy of a red blood cell automatic counting result is improved.

Description

technical field [0001] The invention relates to the technical field of digital image processing, in particular to an automatic counting method for cohesive red blood cells based on hyperspectral imaging. Background technique [0002] As the most common type of blood cell, red blood cells are responsible for oxygen transport and immune function. Red blood cell count is an important indicator of blood routine examination and has important reference value in disease prevention and diagnosis. At present, the automatic blood cell analyzer has effectively improved the analysis rate, but the analysis results have a high false negative rate, and some samples still need to be re-examined by the inspector to reduce the rate of missed diagnosis and misdiagnosis. Microscopic examination of blood smear, as the gold standard for clinically judging the pathological changes of blood cells, is still an essential analysis method. The traditional method of blood cell classification and recog...

AI Technical Summary

Problems solved by technology

However, due to the diversity of cell morphology, cell adhesion, and the interference of certain components, it has not been possible to find a method that can meet the clinical requirements for any cell image processing accuracy.

Images