Authors: Can Li, Xiaoqian Jiang, Kai Zhang

Published on: April 05, 2023

Impact Score: 8.2

Arxiv code: Arxiv:2304.02780

Summary

• What is new: A novel deep-learning model that predicts multiple post-transplant risks with a fairness-achieving algorithm to ensure prediction equity.
• Why this is important: Challenges in liver transplant include finding best donor-patient matches and ensuring transplant equity among subpopulations.
• What the research proposes: A multi-task deep neural network that predicts five post-transplant risks and a fairness-achieving algorithm for equal predictions across subpopulations.
• Results: Reduced task discrepancy by 39% and achieved substantial fairness among sensitive attributes.

Technical Details

Technological frameworks used: Deep-learning, multi-task learning

Models used: Deep neural networks

Data used: Electronic health records of 160,360 liver transplant patients from U.S. liver transplant records (1987-2018)

Potential Impact

Healthcare providers, insurance companies, and liver transplant technology developers could benefit or need to adapt.

Want to implement this idea in a business?

We have generated a startup concept here: TransplantAI.