LiteMIL: a computationally efficient cross-attention multiple instance learning for cancer subtyping on whole-slide images
- 1 January 2026
- journal article
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 13 (01) , 1-14
Abstract
PurposeAccurate cancer subtyping is essential for precision medicine but challenged by the computational demands of gigapixel whole-slide images (WSIs). Although transformer-based multiple instance learning (MIL) methods achieve strong performance, their quadratic complexity limits clinical deployment. We introduce LiteMIL, a computationally efficient cross-attention MIL, optimized for WSIs classification.ApproachLiteMIL employs a single learnable query with multi-head cross-attention for bag-level aggregation from extracted features. We evaluated LiteMIL against five baselines (mean/max pooling, ABMIL, MAD-MIL, and TransMIL) on four TCGA datasets (breast: n=875, kidney: n=906, lung: n=958, and TUPAC16: n=821) using nested cross-validation with patient-level splitting. Systematic ablation studies evaluated multi-query variants, attention heads, dropout rates, and architectural components.ResultsLiteMIL achieved competitive accuracy (average 83.5%), matching TransMIL, while offering substantial efficiency gains: 4.8× fewer parameters (560K versus 2.67M), 2.9× faster inference (1.6s versus 4.6s per fold), and 6.7× lower Graphics Processing Unit (GPU) memory usage (1.15 GB versus 7.77 GB). LiteMIL excelled on lung (86.3% versus 85.0%), TUPAC16 (72% versus 71.4%), and matched kidney performance (89.9% versus 89.7%). Ablation studies revealed task-dependent multi-query performance benefits: Q=4 versus Q=1 improved morphologically heterogeneous tasks (breast/lung +1.3% each, p<0.05) but degraded on grading tasks (TUPAC16: −1.6%), validating single-query optimality for focused attention scenarios.ConclusionsLiteMIL provides a resource-efficient solution for WSI classification. The cross-attention architecture matches complex transformer performance while enabling deployment on consumer GPUs. Task-dependent design insights, single query for sparse discriminating features, multi-query for heterogeneous patterns, guide practical implementation. The architecture’s efficiency, combined with compact features, makes LiteMIL suitable for clinical integration in settings with limited computational infrastructure.Keywords
This publication has 14 references indexed in Scilit:
- Towards a general-purpose foundation model for computational pathologyNature Medicine, 2024
- Multiple instance learning for digital pathology: A review of the state-of-the-art, limitations & future potentialComputerized Medical Imaging and Graphics, 2024
- Domain-Specific Pre-training Improves Confidence in Whole Slide Image ClassificationAnnual International Conference of the IEEE Engineering in Medicine and Biology Society, 2023
- Robustness Fine-Tuning Deep Learning Model for Cancers Diagnosis Based on Histopathology Image AnalysisDiagnostics, 2023
- A self-supervised contrastive learning approach for whole slide image representation in digital pathologyJournal of Pathology Informatics, 2022
- Dual-stream Multiple Instance Learning Network for Whole Slide Image Classification with Self-supervised Contrastive LearningIEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2021
- Data-efficient and weakly supervised computational pathology on whole-slide imagesNature Biomedical Engineering, 2021
- Deep neural network models for computational histopathology: A surveyMedical Image Analysis, 2021
- Pan-cancer diagnostic consensus through searching archival histopathology images using artificial intelligencenpj Digital Medicine, 2020
- Automated classification of histopathology images using transfer learningArtificial Intelligence in Medicine, 2019