Medical World Model: Generative Simulation of Tumor Evolution for Treatment Planning

2025-06-02

AI Summary

This study addresses two critical clinical challenges: the difficulty of predicting dynamic tumor evolution and the lack of computationally tractable models for optimizing individualized treatment regimens. To this end, we propose the first Medical World Model (MeWM), which jointly integrates a vision–language policy module with a tumor generative dynamics module. By incorporating inverse dynamics modeling with survival analysis, MeWM enables causal inference and long-term forecasting of tumor state evolution under diverse therapeutic interventions. Our framework establishes an end-to-end “imaging–intervention–prognosis” modeling paradigm, supporting both treatment planning and quantitative evaluation. In blinded radiologist evaluations, MeWM achieves state-of-the-art specificity. In personalized transarterial chemoembolization (TACE) regimen recommendation, it attains a 13% higher F1-score than a medical-domain fine-tuned GPT model, markedly enhancing clinical decision support capability.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Computer Vision: Multi-modal Vision
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Providing effective treatment and making informed clinical decisions are essential goals of modern medicine and clinical care. We are interested in simulating disease dynamics for clinical decision-making, leveraging recent advances in large generative models. To this end, we introduce the Medical World Model (MeWM), the first world model in medicine that visually predicts future disease states based on clinical decisions. MeWM comprises (i) vision-language models to serve as policy models, and (ii) tumor generative models as dynamics models. The policy model generates action plans, such as clinical treatments, while the dynamics model simulates tumor progression or regression under given treatment conditions. Building on this, we propose the inverse dynamics model that applies survival analysis to the simulated post-treatment tumor, enabling the evaluation of treatment efficacy and the selection of the optimal clinical action plan. As a result, the proposed MeWM simulates disease dynamics by synthesizing post-treatment tumors, with state-of-the-art specificity in Turing tests evaluated by radiologists. Simultaneously, its inverse dynamics model outperforms medical-specialized GPTs in optimizing individualized treatment protocols across all metrics. Notably, MeWM improves clinical decision-making for interventional physicians, boosting F1-score in selecting the optimal TACE protocol by 13%, paving the way for future integration of medical world models as the second readers.

Problem

• Simulating tumor evolution for personalized treatment planning
• Evaluating treatment efficacy via generative tumor progression models
• Optimizing clinical decisions using AI-driven disease dynamics simulation

Innovation

• Generative simulation of tumor evolution dynamics
• Vision-language models for clinical action plans
• Inverse dynamics model for treatment optimization