3D MR and TRUS image segmentation method for prostate system puncture

Abstract

The invention provides a 3D MR and TRUS image segmentation method for prostate system puncture. The method comprises the following steps: respectively obtaining prostate 3D MR and TRUS image data sets; performing preprocessing and data enhancement on the data set to obtain a training data set; constructing a segmentation generator network model, wherein the segmentation generator network comprises an encoder, a decoder, a jump connection and a Bottleneck module; constructing a discriminator network model, wherein the discriminator network is used for distinguishing the output result of the segmentation generator network and the Ground truth, and the discrimination result is used for training of the segmentation generator network; randomly dividing the training data set into two parts, namely a training set and a verification set, and training the segmentation generator network by adopting the training data set; and establishing an evaluation index of the segmentation effect, and evaluating the model effect through an ablation experiment and a contrast experiment. The invention provides a 3D MR and TRUS image segmentation method for prostate system puncture based on a deep learning method, wherein real-time accurate segmentation of 3D MR and TRUS images is realized.

Description

technical field [0001] The invention belongs to the technical field of deep learning image processing, in particular to medical image segmentation technology, in particular to a 3D MR and TRUS image segmentation method for prostate system puncture. Background technique [0002] Prostate cancer ranks second in the world's male cancer mortality rate, and the focus is on early prevention and diagnosis. At present, the mainstream diagnostic method for prostate cancer is prostate specific antigen (Prostate Specific Antigen, PSA) detection, followed by biopsy. Due to the poor specificity and low sensitivity of PSA, needle biopsy is clinically used as the gold standard for the diagnosis of prostate cancer. In the field of prostate interventional diagnosis and treatment, medical image-guided diagnosis and treatment methods occupy the mainstream position. The visualization of prostate anatomy and lesions mainly relies on magnetic resonance (Magnetic Resonance, MR) imaging or transrec...

AI Technical Summary

Problems solved by technology

Traditional automatic segmentation methods are mainly divided into contour-based and shape-based methods, region-based methods, and hybrid methods. Active shape models, active appearance models, level sets, and graph cut methods are widely used. These methods are real-time and versatile. Chemical properties are relatively poor

Images