Robust Brain Tumor Segmentation with Incomplete MRI Modalities Using Hölder Divergence and Mutual Information-Enhanced Knowledge Transfer

2025-07-01

AI Summary

Missing modalities in multi-modal MRI—due to poor image quality, inconsistent acquisition protocols, patient contraindications, or cost constraints—severely degrade brain tumor segmentation performance. Method: We propose a robust single-modal parallel segmentation framework that achieves high-accuracy segmentation using only one input MRI modality (e.g., T1, T2, FLAIR, or T1c). It incorporates dynamic parameter adaptation and modality-adaptive feature compensation to mitigate modality-specific information loss. Crucially, we introduce a novel knowledge transfer strategy jointly regularized by Hölder divergence and mutual information, preserving modality specificity while enhancing cross-modal semantic consistency. Results: Evaluated on BraTS 2018 and 2020 benchmarks under diverse modality-missing scenarios, our method consistently outperforms state-of-the-art approaches, achieving Dice score improvements of 2.1–4.7 percentage points. It demonstrates superior robustness and generalization across heterogeneous clinical settings.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Multimodal MRI provides critical complementary information for accurate brain tumor segmentation. However, conventional methods struggle when certain modalities are missing due to issues such as image quality, protocol inconsistencies, patient allergies, or financial constraints. To address this, we propose a robust single-modality parallel processing framework that achieves high segmentation accuracy even with incomplete modalities. Leveraging Holder divergence and mutual information, our model maintains modality-specific features while dynamically adjusting network parameters based on the available inputs. By using these divergence- and information-based loss functions, the framework effectively quantifies discrepancies between predictions and ground-truth labels, resulting in consistently accurate segmentation. Extensive evaluations on the BraTS 2018 and BraTS 2020 datasets demonstrate superior performance over existing methods in handling missing modalities.

Problem

• Robust brain tumor segmentation with missing MRI modalities
• Enhancing knowledge transfer using divergence and mutual information
• Dynamic network adjustment for incomplete input modalities

Innovation

• Single-modality parallel processing framework
• Holder divergence and mutual information
• Dynamic network parameter adjustment