Cirrhotic portal hypertension non-invasive evaluation method based on 3D multi-channel convolutional neural network

Abstract

The invention relates to the field of non-invasive measurement, discloses a cirrhotic portal hypertension non-invasive evaluation method based on a 3D multi-channel convolutional neural network, and solves the problems that in the prior art, an invasive measurement method is high in risk, high in cost and high in operation difficulty, and the accuracy of a current clinical non-invasive predictionmodel is still affected by various interference factors. The method comprises the following steps: S1, acquiring a CTA layer sequence image of a patient with liver cirrhosis and portal hypertension ina portal period; S2, preprocessing the CTA layer sequence image; S3, performing segmentation based on the preprocessed CTA layer sequence image to obtain three-dimensional images of the liver, the spleen and the portal vein system; S4, extracting the characteristics of the three-dimensional images of the liver, spleen and portal system by adopting a multi-channel 3D convolutional neural network,and training by taking a measured HVPG grading value of the patient with liver cirrhosis portal hypertension as output to obtain a stable portal hypertension prediction model; and S5, evaluating the portal pressure of the patient through the stable portal high-pressure prediction model.

Description

technical field [0001] The invention relates to the field of non-invasive measurement, in particular to a non-invasive evaluation method for liver cirrhosis and portal hypertension based on a 3D multi-channel convolutional neural network. Background technique [0002] Liver cirrhosis is the end-stage manifestation of various chronic liver diseases, and portal hypertension is one of the important manifestations of liver cirrhosis. Accurate assessment of portal hypertension is of great significance for risk stratification and individualized treatment of liver cirrhosis. Hepatic Venous Pressure Gradient (HVPG) is the gold standard for risk stratification of portal hypertension. However, as an invasive measurement method, HVPG suffers from low patient acceptance and high examination costs when there are no serious complications in the early stage of portal hypertension, and its clinical application in my country and other developing countries is greatly limited. Therefore, fin...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to propose a non-invasive evaluation method for cirrhosis and portal hypertension based on 3D multi-channel convolutional neural network, which solves the problem of high risk, high cost and difficult operation of the invasive measurement method in the traditional technology. The accuracy of clinical non-invasive prediction models is still affected by various interference factors

Images