Missing-Modality-Aware Graph Neural Network for Cancer Classification

2025-06-28

AI Summary

To address the challenge of diverse and combinatorially explosive missingness patterns across modalities in multimodal biomedical data, this paper proposes MAGNET—a novel multimodal fusion framework. First, it introduces a missingness-aware patient–modality multi-head attention mechanism to adaptively integrate partially available modality-specific features. Second, it constructs a dynamic patient graph based on similarity of missingness patterns and employs graph neural networks for end-to-end learning. Crucially, MAGNET’s computational complexity scales linearly with the number of modalities, effectively mitigating the exponential growth of missingness pattern space. Evaluated on three real-world cancer multi-omics datasets, MAGNET consistently outperforms state-of-the-art multimodal fusion methods in both classification accuracy and robustness to missing modalities. The source code and datasets are publicly available.

Technology Category

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

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

A key challenge in learning from multimodal biological data is missing modalities, where all data from some modalities are missing for some patients. Current fusion methods address this by excluding patients with missing modalities, imputing missing modalities, or making predictions directly with partial modalities. However, they often struggle with diverse missing-modality patterns and the exponential growth of the number of such patterns as the number of modalities increases. To address these limitations, we propose MAGNET (Missing-modality-Aware Graph neural NETwork) for direct prediction with partial modalities, which introduces a patient-modality multi-head attention mechanism to fuse lower-dimensional modality embeddings based on their importance and missingness. MAGNET's complexity increases linearly with the number of modalities while adapting to missing-pattern variability. To generate predictions, MAGNET further constructs a patient graph with fused multimodal embeddings as node features and the connectivity determined by the modality missingness, followed by a conventional graph neural network. Experiments on three public multiomics datasets for cancer classification, with real-world instead of artificial missingness, show that MAGNET outperforms the state-of-the-art fusion methods. The data and code are available at https://github.com/SinaTabakhi/MAGNET.

Problem

• Handles missing modalities in multimodal biological data
• Addresses diverse missing-modality patterns efficiently
• Improves cancer classification with partial modalities

Innovation

• Patient-modality multi-head attention mechanism
• Linearly scalable with modality count
• Graph neural network with missingness-aware connectivity