Research on Multi-Task Fusion Deep Learning Model for Classification of Hepatic Cystic Hydatidosis

Research must combine AI-assisted medical treatment with improving the visit rate to correctly diagnose and treat hepatic cystic echinococcosis (HCE).Diagnosing hepatic cystic echinococcosis (HCE) largely relies on clinical experience. With the development of imaging and artificial intelligence technologies, constructing deep learning models for classifying HCE can improve diagnostic efficiency. This study aims to improve the diagnostic accuracy of hepatic cystic echinococcosis using a multi-task ensemble deep learning method for classification. A total of 6854 HCE ultrasound images were obtained from patients in this study. We have developed a multi task integrated deep learning method that extracts deep features with a dual branch structure, performs feature fusion and complementarity, and ultimately performs subtype classification. Compare and analyze the experimental results with classical deep learning models. This study investigated and discussed the performance of various deep learning models in the classification of hepatic hydatid cysts, and analyzed the results of CT and ultrasound images from a multitasking perspective.