Abstract:e16501 Background: Survival prediction of cancer patients has always been an challenging problem.Tumor microenvironment (TME) Analyzation based on whole-slide-images (WSIs) has provide an effective perspective for survival prediction. However, most existing TME analyzation based on cell segmentation or classification relies heavily on labor-intensive cell-level annotations of pathologists. Furthermore, except for each individual cell or local pathological feature, survival prediction also involves local-level pathological features interactions in tumor microenvironments. This requires context-awareness based on histological features to fully infer the patient's survival risk. Therefore, we explored a model based on graph convolutional neural networks (GCNN) to perform survival prediction of cancer patients using WSIs. Methods: We utilized WSIs collected from The Cancer Genome Atlas (TCGA) to develop a graph convolutional neural network for survival prediction of cancer patients. The model leverages the advantages of graph structures to autonomously learn the histopathological contextual features in WSIs, and therefore can incorporate additional and crucial tumor microenvironment interaction information while avoiding the labor-intensive annotations. WSIs of two different cancers, KIRC and LUAD, were randomly divided into training, validation, and testing sets in a ratio of 7:1:2. The performance of the constructed model is evaluated using the test set and the results are compared with other state-of-art methods. Results: Our work is compared with other state-of-art weakly supervised learning methods for survival prediction in computational pathology. Abundant experimental results shown that our method outperformed previous methods on these two cancer types (achieving a 2.9% improvement compared to Multiple Instance Learning (MIL) and a 2.6% improvement compared to Attention MIL), with an overall c-index of 0.646. Additionally, we evaluated the interpretability of our model through attention heatmaps of low-risk and high-risk patients. Conclusions: We have developed a GCNN based model, combined with attention mechanisms, to learn features of heterogeneous tumor microenvironments and their contextual information from memory-efficient representations of highly correlated image patches for cancer patients survival prediction. This model is applicable to any weakly supervised learning task in computational pathology that involves slide-level or patient-level labels, making it an effective supplementary diagnostic tool for oncologists and pathologists.

Survival Analysis of High-Dimensional Data With Graph Convolutional Networks and Geometric Graphs

Generating Hypergraph-Based High-Order Representations of Whole-Slide Histopathological Images for Survival Prediction

A novel interactive deep cascade spectral graph convolutional network with multi-relational graphs for disease prediction

Mixed Geometry Message and Trainable Convolutional Attention Network for Knowledge Graph Completion

Graph Learning in Low Dimensional Space for Graph Convolutional Networks

FGCNSurv: dually fused graph convolutional network for multi-omics survival prediction

A Cancer Survival Prediction Method Based on Graph Convolutional Network

Prediction and interpretation of cancer survival using graph convolution neural networks

Annotations-free Survival Prediction with WSIs Using Graph Convolutional Neural Networks.

A novel complex-high-order graph convolutional network paradigm: ChyGCN

Survival Prediction Across Diverse Cancer Types Using Neural Networks

Unified GCNs: Towards Connecting GCNs with CNNs

High-order graph attention network

An Overview of Disease Prediction based on Graph Convolutional Neural Network

A Survey of Geometric Graph Neural Networks: Data Structures, Models and Applications

Scattering GCN: Overcoming Oversmoothness in Graph Convolutional Networks

Improving Cancer Survival Prediction via Graph Convolutional Neural Network Learning on Protein-Protein Interaction Networks

Graph convolutional networks of reconstructed graph structure with constrained Laplacian rank

HPCGCN: A Predictive Framework on High Performance Computing Cluster Log Data Using Graph Convolutional Networks

Attention-Based Deep Recurrent Model for Survival Prediction

Global Minima, Recoverability Thresholds, and Higher-Order Structure in GNNS