Beyond Accuracy: Interpretable Multimodal Deep Learning for Glioma Tumor Grade Classification

DOI

10.2991/978-94-6239-756-9_26How to use a DOI?

Keywords

Glioma Grade classification; Medical imaging; Multimodal Integration; Molecular data; Clinical Data and Explainable AI

Abstract

The most common and dangerous brain tumor is glioma and have a unfavourable outcome, significant morbidity, and high mortality. Clinical decision-making depends on accurate and timely glioma grading. While recent DL models have achieved exceptional results in the classification of medical images, their non-interpretable nature hinders the translations into a clinical setting. This paper proposes an interpretable, multimodal Deep Learning (DL) framework that combines MRI features with clinical and molecular features for robust glioma-grade classification. EfficientNetB4 achieves 99.4% accuracy for MRI-based classification, while XGBoost achieves 92% using molecular and clinical biomarkers. Accuracy further improves to 99.8% with a weighted late-fusion strategy and surpasses the single-modality approaches. Grad-CAM visualizations enhance the interpretability by highlighting the tumor-relevant regions in MRI scans. The proposed framework offers improved diagnostic accuracy and fosters clinical trust due to explainability.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Kavita Jain
AU  - Deepali Vora
AU  - Abderrahim Benslimane
PY  - 2026
DA  - 2026/08/31
TI  - Beyond Accuracy: Interpretable Multimodal Deep Learning for Glioma Tumor Grade Classification
BT  - Proceedings of the Conference on Bioengineering for Global Health (SYMRESEARCH 2.0 2025)
PB  - Atlantis Press
SP  - 371
EP  - 394
SN  - 2468-5747
UR  - https://doi.org/10.2991/978-94-6239-756-9_26
DO  - 10.2991/978-94-6239-756-9_26
ID  - Jain2026
ER  -