Conditioned world modeling improves breast cancer treatment prediction
CoRe-WM: Conditioned Response-State World Modeling for Longitudinal Treatment-Response Prediction in Breast DCE-MRI
Computer Vision and Pattern Recognition
Summary
Predicting how breast cancer patients respond to chemotherapy is important but hard because responses change over time. The authors treat this prediction as a ‘world model’ problem, learning how a patient’s tumor state changes across multiple imaging scans. Their method, CoRe-WM, takes into account each patient’s unique clinical profile and treatment to better anticipate changes in tumor appearance. They show that this approach predicts the likelihood of complete tumor response better than previous methods and works well even on different patient groups.
What this means in practice
- •For radiology teams: Use a model that incorporates patient profiles and treatment history to improve predictions of breast cancer response from sequential MRI images during therapy.
- •For oncology data scientists: Build treatment response models that separate tumor morphology changes from imaging appearance for better outcome predictions in cancer clinical trials.
Authors
Lin Zhao, Bowen Chen, Qiaohui Gao, Shaowen Wan, Xiang Li, Wei Liu, Tianming Liu
Abstract
Pathologic complete response (pCR) after neoadjuvant chemotherapy is a key prognostic indicator in aggressive breast cancer, yet existing prediction approaches either rely on pretreatment imaging alone, missing the patient's evolving response, or optimize multi-visit classifiers under direct outcome supervision, coupling representation learning with the clinical endpoint. We formulate longitudinal pCR prediction as a world-modeling problem and propose Conditioned Response-state World Modeling (CoRe-WM), which learns a latent state-transition model that predicts the next imaging state from the observed sequence. The transition function is conditioned on patient clinical profile and assigned therapy through additive shift and feature-wise modulation, enabling patient- and treatment-specific dynamics. A Factorized Response State pathway further decouples treatment-response dynamics from imaging appearance by encoding the structural prior that response manifests primarily in lesion morphology. After pretraining, all parameters are frozen and a linear readout predicts pCR at each clinical decision point. We evaluate CoRe-WM on I-SPY2 under five-fold cross-validation and assess cross-cohort transfer on I-SPY1. CoRe-WM achieves the highest discrimination among all evaluated methods at every prospective decision point, and demonstrates robust cross-cohort transfer. Ablation studies identify patient-profile conditioning and the factorized response pathway as the principal contributors. These results establish world-model-based representation learning as a promising framework for outcome prediction in oncology and beyond.