Lightweight Transformer-based knowledge distillation framework for high-dimensional spatiotemporal radiomics in breast cancer risk prediction.

Breast cancer exhibits significant spatiotemporal heterogeneity. Traditional radiomics approaches usually rely on low-temporal-resolution imaging and discrete image phases, failing to capture the rapid and continuous kinetic evolution within the tumor microenvironment. Although ultrafast dynamic contrast-enhanced magnetic resonance imaging captures precise contrast agent permeation, current methodologies lack systematic, dedicated computational frameworks to extract these high-dimensional dynamic features. To bridge this gap, this study developed and validated a robust two-stage spatiotemporal radiomics framework for high-dimensional feature selection and predictive modeling. First, a novel gradient-dynamics-based feature selection algorithm was designed to robustly identify highly discriminative kinetic trajectory patterns and overcome the dimensionality curse of time-series data. Second, a lightweight Transformer-based predictive network incorporating a self-attention knowledge distillation mechanism was deployed to capture deep representational knowledge and enhance the differentiation between benign and malignant lesions. Experimental results demonstrated that the proposed framework outperformed traditional baseline models using delta features or pharmacokinetic parameters, achieving an area under the curve of 0.959 ± 0.022 and an accuracy of 92%, an 8% improvement over conventional temporal radiomics models (P < 0.05). Furthermore, through the visualization of attention heat maps, the model provides highly interpretable evidence of the critical enhancement phases and spatiotemporal evolutionary patterns that drive model decisions. This framework introduces a viable, visually interpretable, and quantitative tool for characterizing tumor heterogeneity, demonstrating its potential to assist in personalized clinical decision-making for breast cancer.