Liver tumor MVI prediction method based on triple network

Abstract

The invention provides a liver tumor MVI prediction method based on a triple network. The method comprises the following steps: dividing liver MRI image data into a training set and a test set; selecting three samples from the training set to form a triple sample; constructing a triple network model, wherein the triple network model is composed of three paths of convolutional neural networks with the same structure, and the triple network model receives one triple sample as input each time; converting an input triple sample into a vector in an embedded layer space through the triple network model, calculating triple loss, calculating cross entropy loss by using a convolutional neural network in the triple network model, and fusing the triple loss and the cross entropy loss to serve as a target loss function for training the triple network model; and training the triple network model by using the training set to obtain an optimal triple network model, and performing liver tumor MVI prediction on the liver MRI image data in the test set by using the trained triple network model.

Description

technical field [0001] The invention relates to the technical field of medical image classification, in particular to a triple network-based MVI prediction method for liver tumors. Background technique [0002] Hepatocellular carcinoma (Hepatocellular Carcinoma, HCC) accounts for about 80% of the total liver cancer cases, and is prone to recurrence and distant metastasis, leading to poor prognosis. Microvascular Invasion (MVI) is a risk factor affecting prognosis, which can indicate the recurrence of liver cancer, and is a common pattern of vascular invasion in HCC. Successful preoperative MVI prediction is crucial to determine the treatment options for HCC patients and can significantly improve patient outcomes. The current gold standard for the diagnosis of MVI is obtained through pathological analysis of intraoperative specimens, which has no obvious guiding significance for the preoperative diagnosis of HCC patients. In recent years, non-invasive magnetic resonance ima...

AI Technical Summary

Problems solved by technology

The disadvantage of radiomics-based methods is that the definition and extraction of image features are greatly affected by human factors, and they are subject to some constraints in clinical applications.

At present, the research on MVI prediction of HCC patients using deep learning has some preliminary results, but most of the current methods do not consider the similarity between MRI images of HCC patients or the fine-grained features of the images.

Images