Liver-vein-type bcs drug balloon treatment efficacy prediction method based on ai image analysis
By integrating multimodal imaging and clinical data into a multi-class logistic regression model, we have achieved accurate prediction of restenosis risk after hepatic vein BCS drug balloon therapy. This solves the problems of low prediction accuracy and poor interpretability in existing technologies, and provides an efficient risk assessment tool to support personalized treatment plans.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- THE AFFILIATED HOSPITAL OF XUZHOU MEDICAL UNIV
- Filing Date
- 2026-04-24
- Publication Date
- 2026-07-17
AI Technical Summary
In the current technology, there is a lack of accurate predictive means for evaluating the efficacy of drug balloon therapy for hepatic venous Budd-Chiari syndrome (BCS). Traditional methods are highly subjective and have low predictive accuracy. Furthermore, existing models have many parameters, are difficult to train, and have poor interpretability, which cannot meet clinical needs.
By integrating multimodal medical imaging data and structured clinical data, and employing an improved multi-class logistic regression model, imaging features and clinical features are extracted to generate a unified feature vector, enabling accurate prediction of restenosis risk after drug-eluting balloon treatment in patients with hepatic vein type BCS.
It achieves accurate classification and prediction of restenosis risk, with a classification accuracy of 94.33%, which is significantly better than CNN models. The misclassification rate of each risk level is low, the model is simple and easy to interpret, and it is convenient for clinical application. It provides objective evidence for efficacy evaluation, reduces the occurrence of restenosis, and improves patient prognosis.
Smart Images
Figure CN122417409A_ABST