Search papers, labs, and topics across Lattice.
The paper introduces Multiple Instance Learning on Precomputed Features (MIL-PF), a framework for mammography classification that leverages frozen foundation model encoders and a lightweight MIL head. By precomputing image representations, MIL-PF avoids computationally expensive end-to-end fine-tuning, enabling efficient adaptation to high-resolution medical images with limited annotations. The approach achieves state-of-the-art classification performance at clinical scale while significantly reducing training complexity.
Achieve state-of-the-art mammography classification with a lightweight, efficient method that trains only 40k parameters by freezing foundation model encoders.
Modern foundation models provide highly expressive visual representations, yet adapting them to high-resolution medical imaging remains challenging due to limited annotations and weak supervision. Mammography, in particular, is characterized by large images, variable multi-view studies and predominantly breast-level labels, making end-to-end fine-tuning computationally expensive and often impractical. We propose Multiple Instance Learning on Precomputed Features (MIL-PF), a scalable framework that combines frozen foundation encoders with a lightweight MIL head for mammography classification. By precomputing the semantic representations and training only a small task-specific aggregation module (40k parameters), the method enables efficient experimentation and adaptation without retraining large backbones. The architecture explicitly models the global tissue context and the sparse local lesion signals through attention-based aggregation. MIL-PF achieves state-of-the-art classification performance at clinical scale while substantially reducing training complexity. We release the code for full reproducibility.
