Project duration:
Feb 2022 - Feb 2023
PIONEERS IN HEALTHCARE INNOVATION FUND 2021
Data-driven multimodal decision support for the prediction of complications following esophageal surgery
Project summary:
Treatment of esophageal cancer requires major surgery. Postoperative complications such as anastomotic leakage and pneumonia occur frequently (50%) and are often life-threatening (mortality 11%). Early recognition of a complication is crucial for speedy recovery. An attending physician still must base decisions on experience, gained by interpretation of bedside information, vital parameters, and diagnostics. Decision support with machine learning can forge a breakthrough, allowing a faster and more adequate response to emerging complications. Therefore, we investigate whether simultaneous use of multiple data streams (vitals, laboratory tests, chest images, health status, pain perception and mobility) leads to a better prediction of complications.
In this research project, we focus on the question:
‘How to design, implement and evaluate a multimodal data-driven decision support system, including explanation functionality, for the prediction of the recovery process after esophageal surgery?’ Closely related to this question are the issues of (1) early and reliable detection of complications such as anastomotic leakage and pneumonia, and (2) the determination of early yet safe discharge of a patient, i.e., without developing complications at home that require re-admission.
Project Participant:
Funder:
