Research Project

Survival prediction, one very important task in prognosis, now meets many deep learning approaches. Several deep survival learning models are proposed for predicting the surivals of cancer patients from the provided multi-modal data of patients.

Technological advances create a great opportunity to provide multi-view data for patients including 3D/4D radiological imaging, High resolutional pathological slides, gene expression, etc. For example in imaging data, traditional surival models mainly rely on explicitly-designed handcrafted features. This process highly relies on domain expert and human intervention which could introduce human bias and limit the use on large scale data.

Deep learning have been successfully used in many medical imaging problem including detection, classification, segmentation. Survival prediction describes how patients will survival in the future. Depend on the event, it also includes the predication of Overall Survival (OS), Disease Free Survival (DFS) or other prognostic factor prediction. One of my research work is survival prediction using various imaging modality including CT, pathology and genomics. The following show papers that I authored or co-authored in recent years.

CT

• Improved Prognostic Prediction of Pancreatic Cancer Using Multi-phase CT by Integrating Neural Distance and Texture-Aware Transformer, In MICCAI 2023.

• Deep learning for fully automated prediction of overall survival in patients undergoing resection for pancreatic cancer: a retrospective multicenter study, Annals of surgery 278 (1), e68-e79, 2023

• DeepPrognosis: Preoperative prediction of pancreatic cancer survival and surgical margin via comprehensive understanding of dynamic contrast-enhanced CT imaging and tumor-vascular contact parsing, Medical image analysis 73, 102150, 2021. (MICCAI-MedIA Special Issue of Best Papers in 2020)

Pathology

• Whole Slide Images based Cancer Survival Prediction using Attention Guided Deep Multiple Instance Learning Networks, Medical Image Analysis, 2020. (Most Cited Articles since 2020)

• Deep Multi-instance Learning for Survival Prediction from Whole Slide Images, In MICCAI 2019.

• Graph CNN for survival analysis on whole slide pathological images, In MICCAI 2018.

• Wsisa: Making survival prediction from whole slide histopathological images, In CVPR 2017.

  • One of earliest work developing weakly-supervised survival model for pathology whole slides.

• Deep convolutional neural network for survival analysis with pathological images, In BIBM 2016.

  • The first work developing deep survival model for pathology images, known as DeepConvSurv.

• Imaging biomarker discovery for lung cancer survival prediction, In MICCAI 2016.

PET

• Deep learning for fully automated prediction of overall survival in patients with oropharyngeal cancer using FDG-PET imaging, Clinical Cancer Research 27 (14), 3948-3959, 2021

Image-Omics

• Deep correlational learning for survival prediction from multi-modality data, In MICCAI 2017.

• Imaging-genetic data mapping for clinical outcome prediction via supervised conditional gaussian graphical model, In ISBI 2016.