Pediatric brain tumor classification using digital histopathology and deep learning: evaluation of SOTA methods on a multi-center Swedish cohort

2024-09-02

arXiv.org

AI Summary

Pediatric brain tumor diagnosis lacks large-scale, expertly annotated histopathological datasets, hindering the development of robust computational pathology tools. Method: We constructed a weakly supervised multiple-instance learning (MIL) framework trained on 540 H&E-stained whole-slide images from six Swedish centers, enabling hierarchical classification into tumor category, family, and type. We systematically evaluated tissue-specific foundation models (UNI, CONCH) paired with MIL aggregation strategies (ABMIL, CLAM). Contribution/Results: UNI combined with ABMIL achieved optimal performance, attaining Matthew’s correlation coefficients of 0.86±0.04, 0.63±0.04, and 0.53±0.05 for category, family, and type classification, respectively—significantly outperforming ResNet50. Class activation heatmaps demonstrated strong clinical interpretability. This study demonstrates the feasibility and robustness of state-of-the-art computational pathology methods in resource-constrained pediatric settings.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Computer Vision: Other Foundations of Computer Vision

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Brain tumors are the most common solid tumors in children and young adults, but the scarcity of large histopathology datasets has limited the application of computational pathology in this group. This study implements two weakly supervised multiple-instance learning (MIL) approaches on patch-features obtained from state-of-the-art histology-specific foundation models to classify pediatric brain tumors in hematoxylin and eosin whole slide images (WSIs) from a multi-center Swedish cohort. WSIs from 540 subjects (age 8.5$pm$4.9 years) diagnosed with brain tumor were gathered from the six Swedish university hospitals. Instance (patch)-level features were obtained from WSIs using three pre-trained feature extractors: ResNet50, UNI and CONCH. Instances were aggregated using attention-based MIL (ABMIL) or clustering-constrained attention MIL (CLAM) for patient-level classification. Models were evaluated on three classification tasks based on the hierarchical classification of pediatric brain tumors: tumor category, family and type. Model generalization was assessed by training on data from two of the centers and testing on data from four other centers. Model interpretability was evaluated through attention-mapping. The highest classification performance was achieved using UNI features and AMBIL aggregation, with Matthew's correlation coefficient of 0.86$pm$0.04, 0.63$pm$0.04, and 0.53$pm$0.05, for tumor category, family and type classification, respectively. When evaluating generalization, models utilizing UNI and CONCH features outperformed those using ResNet50. However, the drop in performance from the in-site to out-of-site testing was similar across feature extractors. These results show the potential of state-of-the-art computational pathology methods in diagnosing pediatric brain tumors at different hierarchical levels with fair generalizability on a multi-center national dataset.

Problem

• Classifying pediatric brain tumors using deep learning on histopathology images
• Evaluating weakly supervised MIL approaches for tumor categorization
• Assessing model generalizability across multi-center datasets

Innovation

• Uses weakly supervised multiple-instance learning (MIL)
• Leverages histology-specific foundation models
• Evaluates on multi-center pediatric brain tumor dataset