ImmunoDiff: A Diffusion Model for Immunotherapy Response Prediction in Lung Cancer

2025-05-29

AI Summary

Predicting immunotherapy response in non-small cell lung cancer (NSCLC) is hindered by the difficulty of modeling dynamic radiological changes over time from serial CT scans. Method: We propose an anatomy-aware temporal CT synthesis framework that integrates lobar and vascular segmentation priors with clinical constraints—including PD-L1 expression and blood biomarkers—via a novel diffusion model. We design a cross-biomodal injection adapter (cbi-Adapter) to ensure consistent multimodal conditioning between imaging and clinical data, and establish an interpretable baseline-to-follow-up CT synthesis paradigm. Contribution/Results: Evaluated on an NSCLC cohort, our method improves balanced accuracy for treatment response prediction by 21.24% and increases the concordance index (c-index) for overall survival prediction by 0.03, significantly outperforming static baseline-only models. This work introduces a new paradigm for personalized immunotherapy efficacy assessment grounded in spatiotemporal radiological dynamics.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Accurately predicting immunotherapy response in Non-Small Cell Lung Cancer (NSCLC) remains a critical unmet need. Existing radiomics and deep learning-based predictive models rely primarily on pre-treatment imaging to predict categorical response outcomes, limiting their ability to capture the complex morphological and textural transformations induced by immunotherapy. This study introduces ImmunoDiff, an anatomy-aware diffusion model designed to synthesize post-treatment CT scans from baseline imaging while incorporating clinically relevant constraints. The proposed framework integrates anatomical priors, specifically lobar and vascular structures, to enhance fidelity in CT synthesis. Additionally, we introduce a novel cbi-Adapter, a conditioning module that ensures pairwise-consistent multimodal integration of imaging and clinical data embeddings, to refine the generative process. Additionally, a clinical variable conditioning mechanism is introduced, leveraging demographic data, blood-based biomarkers, and PD-L1 expression to refine the generative process. Evaluations on an in-house NSCLC cohort treated with immune checkpoint inhibitors demonstrate a 21.24% improvement in balanced accuracy for response prediction and a 0.03 increase in c-index for survival prediction. Code will be released soon.

Problem

• Predict immunotherapy response in NSCLC accurately
• Capture complex morphological changes from baseline imaging
• Integrate clinical data to refine predictive models

Innovation

• Diffusion model synthesizes post-treatment CT scans
• Integrates anatomical priors for enhanced CT fidelity
• cbi-Adapter ensures consistent multimodal data integration