CN-121101470-B - Cancer risk degree identification method and system

CN121101470BCN 121101470 BCN121101470 BCN 121101470BCN-121101470-B

Abstract

The invention provides a method and a system for distinguishing the risk degree of cancer, which relate to the technical field of prostate cancer, and the method comprises the steps of obtaining personal information of a screening object and a prostate specific antigen detection result of the screening object; the method comprises the steps of determining a risk value of screening object antigen detection based on screening object personal information and a prostate specific antigen detection result of a screening object, acquiring a prostate magnetic resonance scanning image of the screening object if the risk value of the screening object antigen detection is larger than a preset threshold value, constructing a graph structure, processing the graph structure based on a graph convolution network to determine prostate cancer analysis information, generating a prostate cancer simulated growth video based on the prostate magnetic resonance scanning image of the screening object and the prostate cancer analysis information by using an anti-network, and determining the risk degree of the prostate cancer based on the prostate cancer simulated growth video.

Inventors

JIAN YONG
HUANG FU
LIU ZHIFENG
WANG FENGHUA
HE GU

Assignees

简勇

Dates

Publication Date: 20260512
Application Date: 20250908

Claims (3)

1. A cancer risk level discrimination system, comprising: A first acquisition module for acquiring personal information of a screening object, and prostate specific antigen examination results of the screening object, the first acquisition module further being configured to: preprocessing the prostate magnetic resonance scanning image of the screened object, wherein the preprocessing comprises denoising, standardization and contrast enhancement; an antigen screening module for determining a risk value for screening an object antigen based on the screening object personal information, a prostate specific antigen screening result of the screening object; The second acquisition module is used for acquiring a prostate magnetic resonance scanning image of the screened object if the risk value of screening the object antigen examination is larger than a preset threshold value; A map structure module for constructing a map structure based on a result of a prostate-specific antigen examination of the screening subject, a magnetic resonance image of the prostate of the screening subject, the map structure comprising two nodes and an edge between the two nodes, the two nodes comprising an antigen examination node, a magnetic resonance examination node, a node characteristic of the antigen examination node comprising the screening subject personal information, the prostate-specific antigen examination of the screening subject, a node characteristic of the magnetic resonance examination node comprising a prostate magnetic resonance image of the screening subject, a characteristic of the edge between the nodes comprising a degree of agreement of the antigen examination and the magnetic resonance examination; the analysis information determining module is used for processing the graph structure based on a graph rolling network to determine the analysis information of the prostate cancer, wherein the input of the graph rolling network is the graph structure, the output of the graph rolling network is the analysis information of the prostate cancer, and the analysis information of the prostate cancer comprises the presence or absence of the prostate cancer, the position, the size, the shape, the density, the edge definition of the tumor, the evidence of invasion of surrounding tissues or organs and the reliability of antigen examination; A generation countermeasure module for generating a prostate cancer simulated growth video using a generation countermeasure network based on the prostate magnetic resonance scan image of the screening subject, the prostate cancer analysis information; a risk level determination module for determining a risk level of prostate cancer based on the simulated growth video of prostate cancer, the risk level determination module further for: The method comprises the steps of processing a simulated growth video of the prostate cancer to determine the prostate cancer risk level based on a risk level determining model, wherein the risk level determining model is a long-term neural network model, the risk level determining model comprises an information change sequence determining layer, a risk information determining layer and a prostate cancer risk level determining layer, the input of the information change sequence determining layer is time sequence data of tumor size, time sequence data of tumor shape, time sequence data of tumor position and time sequence data of tumor density, the output of the information change sequence determining layer is time sequence data of tumor size, time sequence data of tumor shape, time sequence data of tumor position and time sequence data of tumor density, the input of the risk information determining layer is time sequence data of tumor size, time sequence data of tumor shape, time sequence data of tumor position, time sequence data of tumor density, the output of the risk information determining layer is the tumor size, the risk level of tumor shape, the risk level of tumor position, the tumor density, the edge level of tumor, the tumor position and the risk level of the tumor, the risk level of the tumor position is determined by the time sequence data of tumor density, the time sequence data of the tumor position and the risk level, the risk level of the tumor position is the risk level.
2. An electronic device comprising a processor, a memory, and a computer program, wherein the computer program is stored in the memory and configured to be executed by the processor to implement the cancer risk level discrimination system of claim 1.
3. A computer-readable storage medium, on which a computer program is stored, characterized in that the program, when executed by a processor, implements the cancer risk level discrimination system according to claim 1.

Description

Cancer risk degree identification method and system Technical Field The invention relates to the technical field of prostate cancer, in particular to a method and a system for distinguishing the risk degree of cancer. Background Prostate cancer is one of the most common malignant tumors in men, severely threatening the life health of men. The main methods of prostate cancer screening include Prostate Specific Antigen (PSA) screening and digital rectal screening (DRE). However, in existing clinical practice, PSA levels are often relied solely on to determine whether a prostate biopsy is required, which, while simple, is subject to increased accuracy, since elevated PSA levels do not always mean that prostate cancer is present, but may also be caused by other non-malignant factors, such as hyperplasia or inflammation of the prostate, etc. In recent years, multiparameter magnetic resonance imaging (mpMRI) has become an important tool for prostate cancer diagnosis. mpMRI can provide high resolution images of the internal structure of the prostate, helping to identify potentially malignant lesions. Prostate image report and data system (PI-RADS) based on mpMRI is used for grading prostate cancer with great clinical significance through a standardized grading system, so that the accuracy of prostate cancer screening is remarkably improved. The PI-RADS score is, however, primarily dependent on the subjective judgment of the radiologist, which may lead to inconsistencies and errors in the diagnostic grading of prostate cancer. How to accurately evaluate the risk of developing prostate cancer is thus a current challenge. Disclosure of Invention The invention mainly solves the technical problem of accurately evaluating the risk of the occurrence of the prostate cancer. According to a first aspect, the invention provides a cancer risk degree distinguishing method, which comprises the steps of obtaining personal information of a screening object and a prostate specific antigen detection result of the screening object, determining a risk value of the screening object antigen detection based on the personal information of the screening object and the prostate specific antigen detection result of the screening object, obtaining a prostate magnetic resonance scanning image of the screening object if the risk value of the screening object antigen detection is larger than a preset threshold value, constructing a graph structure based on the prostate specific antigen detection result of the screening object and the prostate magnetic resonance scanning image of the screening object, wherein the graph structure comprises two nodes and one side between the two nodes, the two nodes comprise an antigen detection node and a magnetic resonance detection node, node characteristics of the antigen detection node comprise the personal information of the screening object and the prostate specific antigen detection result of the screening object, node characteristics of the magnetic resonance detection node comprise the prostate magnetic resonance scanning image of the screening object, characteristics of sides between the nodes comprise the antigen detection node and the prostate detection node comprise the prostate magnetic resonance image of the screening object, generating a graph structure based on the prostate resonance network growth simulation graph, and determining the cancer growth degree based on the prostate growth simulation graph and the prostate growth simulation graph, and the cancer growth simulation graph structure is made to be determined based on the prostate growth simulation graph. In one possible implementation, the determining the risk level of the prostate cancer based on the video of the simulated growth of the prostate cancer comprises processing the video of the simulated growth of the prostate cancer based on a risk level determination model, wherein the risk level determination model is a long-short-term neural network model. In one possible implementation, the acquiring a plurality of time-sequentially arranged prostate cancer images further includes acquiring a prostate magnetic resonance scan image of the screening subject further includes preprocessing the prostate magnetic resonance scan image of the screening subject, including denoising, normalization, and contrast enhancement. In one possible implementation, the input of the graph rolling network is the graph structure, and the output of the graph rolling network is the prostate cancer analysis information. According to a second aspect, the present invention provides a cancer risk level discrimination system comprising: the first acquisition module is used for acquiring personal information of a screening object and a prostate specific antigen detection result of the screening object; an antigen screening module for determining a risk value for screening an object antigen based on the screening object personal information, a prostate specific antigen