Deep Semi-Supervised Survival Analysis for Predicting Cancer Prognosis

2026-01-28

AI Summary

This study addresses the critical challenge of limited labeled survival data in high-dimensional settings, which severely hampers the performance of deep Cox models. To overcome this limitation, the work introduces deep semi-supervised learning to survival analysis for the first time, proposing the Cox-MT model based on the Mean Teacher framework. Cox-MT jointly leverages both labeled and unlabeled data to train uni- and multimodal neural networks, effectively integrating diverse modalities such as RNA-seq and whole-slide images. Evaluated on four cancer datasets from The Cancer Genome Atlas (TCGA), the method significantly outperforms existing approaches like Cox-nnet. Moreover, its generalization capability improves with increasing amounts of unlabeled data, achieving optimal performance particularly in multimodal configurations.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Data Mining & Knowledge Management: Mining of Visual, Multimedia & Multimodal Data

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• 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

Abstract

The Cox Proportional Hazards (PH) model is widely used in survival analysis. Recently, artificial neural network (ANN)-based Cox-PH models have been developed. However, training these Cox models with high-dimensional features typically requires a substantial number of labeled samples containing information about time-to-event. The limited availability of labeled data for training often constrains the performance of ANN-based Cox models. To address this issue, we employed a deep semi-supervised learning (DSSL) approach to develop single- and multi-modal ANN-based Cox models based on the Mean Teacher (MT) framework, which utilizes both labeled and unlabeled data for training. We applied our model, named Cox-MT, to predict the prognosis of several types of cancer using data from The Cancer Genome Atlas (TCGA). Our single-modal Cox-MT models, utilizing TCGA RNA-seq data or whole slide images, significantly outperformed the existing ANN-based Cox model, Cox-nnet, using the same data set across four types of cancer considered. As the number of unlabeled samples increased, the performance of Cox-MT significantly improved with a given set of labeled data. Furthermore, our multi-modal Cox-MT model demonstrated considerably better performance than the single-modal model. In summary, the Cox-MT model effectively leverages both labeled and unlabeled data to significantly enhance prediction accuracy compared to existing ANN-based Cox models trained solely on labeled data.

Problem

• Survival Analysis
• Cox Proportional Hazards Model
• Semi-Supervised Learning
• Cancer Prognosis
• High-Dimensional Data

Innovation

• Deep Semi-Supervised Learning
• Cox Proportional Hazards Model
• Mean Teacher Framework
• Multi-Modal Survival Analysis
• Cancer Prognosis Prediction