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

TL;DR

The Medical World Model (MeWM) uses generative AI to simulate tumor evolution and optimize treatment plans, enhancing clinical decision-making with state-of-the-art tumor prediction and treatment efficacy evaluation.

Contribution

This paper introduces the first medical world model combining vision-language and tumor generative models for simulating disease dynamics and optimizing treatment strategies.

Findings

Achieves state-of-the-art tumor prediction accuracy in Turing tests.

Outperforms GPT-based models in treatment protocol optimization.

Improves clinical decision-making metrics, such as F1-score by 13%.

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.