Improved Faster RCNN-based microscopic hyperspectral white blood cell detection method

Abstract

The invention belongs to the technical field of medical treatment, and provides an improved Faster RCNN-based microscopic hyperspectral image white blood cell detection method, which is a method for acquiring a blood smear white blood cell detection image by using a hyperspectral microscope and identifying and classifying white blood cells based on the improved Faster RCNN. The method comprises the following steps: acquiring a pseudo-color image and spectral data of a blood smear by using a hyperspectral microscope, labeling different types of white blood cells, and making a data set; improving a traditional Faster RCNN network, namely, Resnet18 is used for replacing VGG16 in an original Faster RCNN network, and is used as a new pseudo-color image feature extraction network; and establishing a spectral data extraction module aiming at blood smear hyperspectral data, carrying out spectral feature extraction by using a one-dimensional convolutional neural network, and on the basis of the improvement, performing white blood cell image feature and spectral feature fusion by using the Faster RCNN network to finally achieve white blood cell identification and classification. Compared with a traditional Faster RCNN network, the method has the advantage that the identification precision and the classification accuracy of the white blood cells are remarkably improved.

Description

technical field [0001] The invention belongs to the field of medical technology, in particular to a microscopic hyperspectral white blood cell detection method based on an improved Faster RCNN. Background technique [0002] White blood cells are an important part of blood, produced by bone marrow and lymphoid tissue, and have the function of fighting viral and bacterial infections. Can be divided into neutrophils, lymphocytes, monocytes, eosinophils, basophils. The traditional manual microscope detection method is too complicated, especially when detecting a large number of samples, the manual classification and counting of white blood cells is more prone to errors. In recent years, corresponding computer vision algorithms and systems have been widely used in the field of automatic classification and detection of blood cells. [0003] Computer vision detection can be divided into two categories at present. The first category is to simply stack traditional algorithm modules...

AI Technical Summary

Problems solved by technology

Although hyperspectral imaging provides rich spectral information on the basis of spatial features, hundreds of narrow continuous bands make the data dimensionality redundant, requiring manual extraction of spectral values ​​of the region of interest (ROI), which is difficult in actual monitoring. extremely cumbersome

Images