Survival Analysis Using Deep Learning with Medical Imaging

• Activity Number: 79 - Contributed Poster Presentations: Lifetime Data Science Section
• Type: Contributed
• Date/Time: Monday, August 3, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Lifetime Data Science Section
• Author(s): Samantha Morrison* and Constantine Gatsonis and Ani Eloyan and Jon Steingrimsson
• Companies: Brown University and Brown University and Brown University and Brown University
• Keywords: Convolutional Neural Networks; Censoring Unbiased Transformations; Time-to-event outcomes

Abstract

There is widespread interest in using deep learning to build prediction models for medical imaging data. These deep learning methods capture the local structure of the image and require no manual feature extraction. Despite the importance of survival analysis for the analysis of medical data, research on deep learning with time-to-event imaging analyses is still under-developed. We provide an overview of deep learning methods for time-to-event outcomes, propose a method of convolutional neural networks with censoring unbiased loss functions that do not rely on the proportional hazards assumption, and compare the method to previously developed Cox proportional likelihood based neural networks through the analysis of a histology dataset of gliomas.