Tissue classification method and device based on cardiovascular IVOCT images

Abstract

The invention relates to a tissue classification method and device based on cardiovascular IVOCT images. The method comprises the following steps: step 1 of acquiring a plurality of labeled IVOCT images; step 2 of establishing an IVOCT image sample set and dividing the IVOCT images sample set into a training sample set and a test sample set; step 3 of constructing a convolutional neural network structure; step 4 of utilizing the training sample set to train the convolutional neural network, to obtain a CNN model; and step 5 of inputting the test sample set into the CNN model, to obtain tissuetype graphs corresponding to different tissues. In the tissue classification method and device based on cardiovascular IVOCT images, the CNN model has two output ends, which display the outline and the internal structure of the tissue respectively, and the technical problem that in the prior art, the resolution of an output image is slightly small, so that the tissue boundary display is not clearafter the loss of some image information is solved.

Description

technical field [0001] The invention belongs to biological tissue imaging technology, in particular to a method and device for tissue classification based on cardiovascular IVOCT images. Background technique [0002] A biopsy is a common medical test in which a pathologist looks at a sample of tissue taken from a subject under a microscope to determine the nature or extent of disease. Typically, tissue needs to be cut into very thin sections and stained before it can be viewed under a microscope. Optical coherence tomography (OCT) is an alternative to non-destructive optical imaging modalities that can provide high-resolution 3D images of tissue from biopsy samples without staining. Optical coherence microscopy (OCM) combines the advantages of OCT and confocal microscopy to provide high-resolution cellular images. [0003] Normal arteries have a uniform layered structure composed of intima, media and adventitia, but when blood vessels are diseased, different types of tissu...

AI Technical Summary

Problems solved by technology

[0003] Normal arteries have a uniform layered structure consisting of intima, media and adventitia, but when blood vessels are diseased, different types of tissues will be contained in the blood vessels. Therefore, it is necessary to classify and detect these different tissues. However, , until now, the detection and classification of these tissues mainly rely on manual work, which is very time-consuming

[0004] A convolutional network for biomedical image segmentation is proposed in the prior art. This method uses a contraction path to capture content, an enlargement path to precisely locate, and two paths to form a U-shape, called U-Net. The final output image resolution of U-Net will be slightly smaller than the original image, resulting in the loss of image information and unable to truly restore the organizational structure

Images