Deep Neural Network for Survival Analysis of End-Stage Kidney Disease

Presentation Type

Poster

Student

Yes

Abstract

We are investigating the use of deep neural networks when performing survival Analysis. Traditionally, the Cox Proportional Hazard (PH) model has been used to find estimates for the parameters that affect the survival time for a given disease. However, the Cox PH model fails to capture the nonlinear behavior in the data. In this paper, we employ a deep survival model approach to model the survival time for a person with end-stage renal disease. The goal of the analysis is to find the effect of various factors on ESKD survival time. Our results indicate that, compared to the Cox PH model, the deep survival model provides more accurate results with Harrell's C-index. Providing preliminary results of deep survival neural networks on USRDS data.

Start Date

2-6-2024 1:00 PM

End Date

2-6-2024 2:00 PM

This document is currently not available here.

Share

COinS
 
Feb 6th, 1:00 PM Feb 6th, 2:00 PM

Deep Neural Network for Survival Analysis of End-Stage Kidney Disease

Volstorff A

We are investigating the use of deep neural networks when performing survival Analysis. Traditionally, the Cox Proportional Hazard (PH) model has been used to find estimates for the parameters that affect the survival time for a given disease. However, the Cox PH model fails to capture the nonlinear behavior in the data. In this paper, we employ a deep survival model approach to model the survival time for a person with end-stage renal disease. The goal of the analysis is to find the effect of various factors on ESKD survival time. Our results indicate that, compared to the Cox PH model, the deep survival model provides more accurate results with Harrell's C-index. Providing preliminary results of deep survival neural networks on USRDS data.