Method for automatically generating target region of MRI benign prostatic hyperplasia based on deep neural network

Abstract

The invention discloses a method for automatically generating target region of MRI benign prostatic hyperplasia based on deep neural network, and the method comprises the following steps: 1), obtaining a nuclear magnetic resonance image of a benign prostatic hyperplasia patient before an operation, and carrying out the slicing and data preprocessing; 2) obtaining physiological contour sub-images of the slice in a middle zone and a peripheral zone of a horizontal plane; 3) completing three-dimensional segmentation of the physiological contour of the prostate, and generating a three-dimensional point cloud model of the physiological structure of the prostate; 4) establishing a virtual artificial potential field based on the prostate central zone slice, and performing enhanced protection on the key direction to obtain the distribution of safety coefficients on the slice so as to obtain a reasonable target area; and 5) performing three-dimensional reconstruction on the prostate slice containing the target area contour to finally obtain a three-dimensional benign prostatic hyperplasia operation target model. By means of the method, three-dimensional reconstruction of the prostate and the BPH target area can be achieved, and improvement of the safety and efficiency of the BPH operation is promoted.

Description

technical field [0001] The invention relates to the technical field of medical image processing, in particular to a method for automatically generating an MRI benign prostatic hyperplasia target area based on a deep neural network. Background technique [0002] Benign Prostate Prostate (BPH) is one of the most common symptomatic diseases in middle-aged and elderly men. As we age, the glands in the central zone of the prostate slowly proliferate, which can restrict the flow of urine and block the urethra. BPH can cause lower urinary tract symptoms (LUTS), including poor urination, frequent urination, sequelae dribbling, incomplete urination and nocturnal urination, etc., which seriously affects the normal life of the elderly. With the aging of human society, the incidence and prevalence of BPH and LUTS are increasing rapidly. BPH affects 8 percent of men in their 30s and 50 percent of men in their 50s, and this number increases to greater than 80 percent as men reach their ...

AI Technical Summary

Problems solved by technology

Meanwhile, the surgical field needs to be easily modified by the surgeon since the accuracy and robustness of the deep learning method is worse than that of the doctor's experience

As of now, there is no method to automatically reconstruct the target region of BPH under MRI images

Images