Image feature acquisition method of prostate MRI three-dimensional image

Abstract

The invention relates to an image feature acquisition method of a prostate MRI three-dimensional image. The method comprises: the prostate T2WI image being subjected to automatic organ segmentation; and mapping the prostate organ region to the registered ADC and DWI image based on a segmentation result to obtain a multi-parameter prostate organ region serving as input of a discrimination model, and obtaining image characteristics by combining the multi-parameter MRI image and a deep learning algorithm. The method is based on a large amount of prostate image data, the prostate organ automatic segmentation model is established, the interference of irrelevant background information on the discrimination model is reduced, the characteristics of the multi-parameter MRI image are fused by usinga deep learning method, high-precision prostate cancer characteristic extraction is realized, and a basis is provided for improving the prostate cancer diagnosis efficiency and accuracy.

Description

technical field [0001] The invention relates to an image feature acquisition method of a prostate MRI three-dimensional image. The acquired image features can be used for auxiliary diagnosis of prostate cancer, and belong to the technical field of computer-aided diagnosis of medical images. Background technique [0002] Prostate cancer is a common malignant tumor in men and the second most common cancer in men worldwide. The incidence of prostate cancer in Chinese men is also increasing day by day. Magnetic resonance imaging (MRI) examination has become an important method for the diagnosis of prostate cancer in hospitals due to its advantages of non-invasive examination, diversified scanning images, and prominent focus signals. MRI imaging includes T2WI (T2-weighted imaging), DWI (diffusion-weighted imaging), ADC ( Surface diffusion coefficient) and other sequences assist radiologists in diagnosis. Among them, the T2WI image provides a clearer tissue contrast, and the DWI ...

AI Technical Summary

Problems solved by technology

[0004] At present, there are some methods for prostate cancer discrimination, and some use deep learning methods to use the entire image image as the input of the model, but some irrelevant tissues around the prostate organ will be included in the MRI scanning process, and this method will vary due to the surrounding organs Differences in non-prostate regions lead to misjudgment; other methods only use T2WI images, ignoring the characteristics of other sequences such as ADC and DWI images of the degree of restricted diffusion of water molecules, which are of great value in the identification of prostate malignancies; and Although some methods use the prostate organ region as the input of the discriminant model, doctors need to manually mark the prostate organ region for each diagnosis, which cannot be automatically segmented according to the original MRI image, which increases the cost and cycle of diagnosis

Images