Blood vessel lumen automatic segmentation method based on deep learning

Abstract

The invention provides a blood vessel lumen automatic segmentation method based on deep learning. An accurate segmentation result can be rapidly obtained in real time. The method comprises the following steps: S1, acquiring an IVUS image; s2, after an original image and the marked template image are obtained, establishing a training set and a test sample set; s3, coarsening the template image marked by the training set to obtain a template image, wherein the current frame, the previous frame and the next frame of the training set and the coarsened template image form a four-channel image; s4,enabling the established deep learning segmentation model to adopt a network structure with residual connection, and inputting the template image marked in the S2 and the four-channel image constructed in the S3 into the model to obtain a trained network; and S5, forming a four-channel image by the previous frame segmentation result and the current frame, the previous frame and the next frame of images of the test sample set, inputting the four-channel image into a network for segmentation, and finally obtaining a lumen-intima interface and a middle-outer membrane interface segmented by the current frame.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an automatic segmentation method of blood vessel lumen based on deep learning. Background technique [0002] Intravascular imaging Intravascular ultrasound (IVUS) uses high-frequency ultrasound to present high-resolution images in blood vessels, which can comprehensively analyze the conditions of blood vessel walls and lumens. It is currently the most commonly used intravascular imaging technology in clinical practice. , which can perform high-precision lumen segmentation on intravascular ultrasound images, determine the lumen-intima interface and the media-adventitia interface, and can obtain detailed lumen and plaque information to help doctors make clinical diagnoses. At the same time, Based on the segmentation results, a high-precision three-dimensional model of blood vessels can be established to help improve the accuracy of the virtual blood flow fraction calculati...

AI Technical Summary

Problems solved by technology

[0005] In view of the above problems, the present invention provides an automatic segmentation method of blood vessel lumen based on deep learning, which can solve the problem that the image cannot be segmented due to speckle noise, image artifacts and partial vascular wall calcification shadows, and can quickly obtain accurate Segmentation results with high accuracy and high degree of automation

Images