AI-Based Detection of In-Treatment Changes from Prostate MR-Linac Images

2026-02-04

AI Summary

This study addresses the detection of subtle, therapy-induced anatomical changes occurring over short intervals (mean 2 days) during prostate radiotherapy using routinely acquired MR-Linac images. To this end, we propose a deep learning-based temporal prediction model that identifies such changes by determining the chronological order of longitudinal image pairs. The model integrates saliency maps, input ablation, and quantitative imaging analysis to localize regions most indicative of change. Evaluated on data from 761 patients, the model achieves excellent performance (F1-FL AUC = 0.99, All-pairs AUC = 0.97), significantly outperforming radiologists. It successfully identifies the prostate, bladder, and pubic symphysis as primary sites of anatomical variation, with findings validated by clinical experts. This work represents the first high-accuracy AI-driven detection of short-interval radiotherapy-induced changes, thereby expanding the utility of MR-Linac in treatment monitoring.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Planning, Routing, and Scheduling: Temporal Planning
• Computer Vision: Multi-modal Vision

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Purpose: To investigate whether routinely acquired longitudinal MR-Linac images can be leveraged to characterize treatment-induced changes during radiotherapy, particularly subtle inter-fraction changes over short intervals (average of 2 days). Materials and Methods: This retrospective study included a series of 0.35T MR-Linac images from 761 patients. An artificial intelligence (deep learning) model was used to characterize treatment-induced changes by predicting the temporal order of paired images. The model was first trained with the images from the first and the last fractions (F1-FL), then with all pairs (All-pairs). Model performance was assessed using quantitative metrics (accuracy and AUC), compared to a radiologist's performance, and qualitative analyses - the saliency map evaluation to investigate affected anatomical regions. Input ablation experiments were performed to identify the anatomical regions altered by radiotherapy. The radiologist conducted an additional task on partial images reconstructed by saliency map regions, reporting observations as well. Quantitative image analysis was conducted to investigate the results from the model and the radiologist. Results: The F1-FL model yielded near-perfect performance (AUC of 0.99), significantly outperforming the radiologist. The All-pairs model yielded an AUC of 0.97. This performance reflects therapy-induced changes, supported by the performance correlation to fraction intervals, ablation tests and expert's interpretation. Primary regions driving the predictions were prostate, bladder, and pubic symphysis. Conclusion: The model accurately predicts temporal order of MR-Linac fractions and detects radiation-induced changes over one or a few days, including prostate and adjacent organ alterations confirmed by experts. This underscores MR-Linac's potential for advanced image analysis beyond image guidance.

Problem

• MR-Linac
• radiotherapy-induced changes
• prostate cancer
• longitudinal imaging
• inter-fraction changes

Innovation

• deep learning
• MR-Linac
• treatment-induced changes
• temporal ordering
• saliency mapping