From Exams to Trajectories: Neural Controlled Differential Equations for Longitudinal Mammography Vision‑Language Pretraining

Longitudinal analysis of a patient’s screening history is fundamental to mammography interpretation. Yet existing deep learning models struggle with the irregular, continuous-time nature of screening data, where patient histories involve unevenly timed multi-view exams paired with evolving textual reports. To address this gap, we introduce Dynamo, a novel multi-modal pretraining dynamic framework for longitudinal mammography with two key innovations to model exam timelines as latent trajectories, i.e., a coarse-fine grained exam-level temporal encoder based on Neural Controlled Differential Equations (Neural CDEs) and a Temporal Visual Question Answering (TVQA) loss for query-driven masked token prediction conditioned on temporal context. We perform comprehensive evaluations on downstream tasks including risk prediction using large-scale datasets (EMBED and CSAW-CC), BI-RADS assessment, and breast density classification. Across all benchmarks, Dynamo achieves overall gains over state-of-the-art vision-only and CLIP-style models, improving both calibration and temporal reasoning. On the EMBED dataset, representative gains include 11.81% relative reduction in Risk Prediction’s Brier Score, 3.59% relative improvement in BI-RADS ϰ, and 2.39% relative improvement in BI-RADS AUC in zero-shot settings. Our code is available at this URL.