Red blood cell detection and classification method and device, computer storage medium and electronic equipment

Abstract

The invention discloses a red blood cell detection and classification method and device, a computer storage medium and electronic equipment, and belongs to the field of blood cell detection. The method comprises the following steps: firstly, carrying out full-view photography on a large number of blood slides, and establishing a slide scanning image group; a training data set and a verification data set are prepared for an image restoration model, an image detection model and a multi-task classification model from a slide scanning image group. And performing model training by using deep learning and convolutional neural network technologies, and obtaining a final image restoration model, an image detection model and a multi-task classification model by continuously performing parameter tuning and error analysis optimization on the model. In the application process, a blood slide to be detected is also subjected to full-view photography, and a slide scanning image is established. And obtaining a restored clear slide scanning image through the image restoration model, and obtaining the position and category of the red blood cells through image detection and a multi-task classification model. According to the invention, erythrocyte detection and classification can be realized rapidly, comprehensively, accurately and objectively.

Description

technical field [0001] The invention relates to the field of blood cell detection, in particular to a red blood cell detection and classification method, device, computer storage medium and electronic equipment. Background technique [0002] At present, the blood test process in the hospital is: blood sample - blood analyzer - push dye machine - manual microscopic examination, and the whole process takes 60 minutes. Manually draw blood from the patient to obtain blood samples; obtain various blood cell counts, white blood cell classification and hemoglobin content through a blood analyzer; dye and mark through a slide machine to obtain slides for manual microscopic examination; finally, professional doctors conduct After manual microscopic examination, the blood cell morphology analysis results obtained by manual analysis, including abnormal red blood cell count, abnormal red blood cell classification, etc. [0003] Existing hematology analyzers (hemocytometers, blood cell ...

AI Technical Summary

Problems solved by technology

[0008] 1. The instrument is not only expensive, but also takes a long time to detect and multi-task classification of single-field red blood cell pictures, has poor real-time performance and low efficiency, and it is difficult to analyze the whole field of view

[0009] 2. The category of red blood cells may change at different stages in the growth process. Manual microscopic examination will be subjective, and the re-examination rate is high

[0010] 3. Red blood cells are rich in shape and can be classified from different angles, but the existing technology analyzes red blood cells from a single angle, which covers less information, resulting in insufficient comprehensive and specific analysis results

[0011] 4. The number of types of red blood cells that can be analyzed by existing instruments is limited, resulting in an objective and comprehensive analysis of red blood cells.

[0012] 5. For the blood slides to be tested, the existing technology manually selects a single-view image with uniform distribution of red blood cells. The work efficiency is low, and it is usually difficult to analyze blood cells in a full view

Images