Search papers, labs, and topics across Lattice.
This paper addresses the challenge of sparse supervision in whole slide image (WSI) analysis by introducing a spatially regularized multiple instance learning (MIL) framework. The method leverages the inherent spatial relationships among patch features as label-independent regularization, learning a shared representation space through joint optimization of feature-induced spatial reconstruction and label-guided classification. Experiments on public datasets demonstrate that this approach significantly improves performance compared to existing state-of-the-art MIL methods.
Spatial relationships between image patches can be exploited as a powerful, label-free regularizer to significantly improve whole slide image analysis, even with limited annotations.
Whole slide images, with their gigapixel-scale panoramas of tissue samples, are pivotal for precise disease diagnosis. However, their analysis is hindered by immense data size and scarce annotations. Existing MIL methods face challenges due to the fundamental imbalance where a single bag-level label must guide the learning of numerous patch-level features. This sparse supervision makes it difficult to reliably identify discriminative patches during training, leading to unstable optimization and suboptimal solutions. We propose a spatially regularized MIL framework that leverages inherent spatial relationships among patch features as label-independent regularization signals. Our approach learns a shared representation space by jointly optimizing feature-induced spatial reconstruction and label-guided classification objectives, enforcing consistency between intrinsic structural patterns and supervisory signals. Experimental results on multiple public datasets demonstrate significant improvements over state-of-the-art methods, offering a promising direction.
