**Design of an Explainable Multimodal Artificial Intelligence Framework for Personalized Prediction of Fear of Cancer Recurrence in Liver Cancer Patients Using CT Imaging, Clinical Data, and Psychological Indicators**
**Design of an Explainable Multimodal Artificial Intelligence Framework for Personalized Prediction of Fear of Cancer Recurrence in Liver Cancer Patients Using CT Imaging, Clinical Data, and Psychological Indicators**
zahra baaji kazemi1 Parvin Mahmoudi2
1) Department of Computer Engineering, Ar.C., Islamic Azad University, Arak, Iran
2) Department of Clinical Psychology Arak Branch, Islamic Azad University, Ashtian, Iran .
IOI Code
:
AAHD-ACZAGD
Publication :
National Conference on Information Technology, Nanotechnology, Artificial Intelligence and Technological Futures Studies(itnaf.ir)
Abstract :
**Abstract ** Fear of Cancer Recurrence (FCR) is a major psychological concern among liver cancer patients, yet existing research has largely focused on psychological questionnaires or clinical outcomes. This article proposes an explainable multimodal AI framework that integrates CT images, clinical data, and psychological indicators to predict individualized FCR risk. Deep learning and radiomics are used to extract imaging features, while machine-learning models such as XGBoost, Random Forest, and neural networks integrate multimodal information. Explainable AI methods such as SHAP are incorporated to identify the key factors influencing predictions. The proposed framework may enable early identification of patients at high risk of FCR and support personalized psychological interventions and clinical decision-making.
Keywords : fear of cancer recurrence liver cancer hepatocellular carcinoma multimodal artificial intelligence machine learning deep learning ct imaging explainable ai psycho oncology personalized medicine