Non-invasive evaluation method of hepatic vein pressure gradient based on multi-modal images and empirical knowledge

Abstract

The invention discloses a non-invasive evaluation method of hepatic vein pressure gradient based on multi-modal images and empirical knowledge. A convolutional neural network is used for feature extraction of multi-modal medical images, and the estimation value of HVPG based on the multi-modal images is obtained. The deep neural network is used for regression analysis of relevant empirical knowledge parameters of the HVPG and the HVPG estimation values based on empirical knowledge are obtained; the HVPG estimation values based on the multi-modal images and empirical knowledge and obtained fromthe steps are fused, and the fused HVPG estimation value is obtained. The convolutional neural network and the deep neural network are trained jointly by an optimization algorithm; after the trainingis completed, the HVPG can be accurately predicted, and the HVPG quantitative estimation value based on the multi-modal images and empirical knowledge is obtained. The complementary information of multi-modal medical images is taken into account, features are supplemented by the corresponding empirical knowledge, and the method is more in line with the medical pertinence.

Description

technical field [0001] The invention relates to the technical field of medical images, in particular to a noninvasive evaluation method for hepatic vein pressure gradient based on multimodal images and empirical knowledge. Background technique [0002] Portal hypertension is one of the most common serious complications of liver cirrhosis. The clinical manifestations are esophageal and gastric varices with rupture and bleeding, ascites, splenomegaly and hypersplenism, etc. As portal pressure increases, the risk of bleeding from esophageal and gastric varices increases, as does the incidence of hepatocellular carcinoma (HCC), the incidence of liver failure after HCC resection, and the risk of associated death . Therefore, accurate quantification and dynamic monitoring of portal hypertension levels are of vital significance to the research on the pathogenesis, diagnosis and treatment of portal hypertension. [0003] The currently recognized "gold standard" for evaluating port...

AI Technical Summary

Problems solved by technology

[0005] 1) Due to the complex changes in liver and spleen morphology, hardness, and hemodynamics secondary to portal hypertension, traditional imaging methods are mostly qualitative diagnoses, which are greatly influenced by subjective experience and lack more accurate quantitative assessment;

[0006] 2) Images of different modalities usually reflect different pathological features, while traditional imaging methods mostly study from the perspective of single-modal imaging technology, and cannot achieve comprehensive multi-dimensional evaluation

Images