Biomedical Engineering, Masterarbeit
Predictive Model for Intraoperative Fluid Management to Minimize Anastomotic Leakage
Predictive Model for Intraoperative Fluid Management to Minimize Anastomotic Leakage
Anastomotic leakage is a critical and potentially life-threatening complication following gastrointestinal surgery. This complication occurs when the connection between two segments of the intestine (anastomosis) fails, leading to the leakage of luminal contents into the abdominal cavity. Intraoperative fluid management plays a crucial role in maintaining stability during surgery. However, both fluid overload and fluid deficit can adversely affect tissue perfusion and healing at the anastomotic site, thereby increasing the risk of leakage. Thus, achieving an optimal balance in fluid administration is essential. Traditional approaches to intraoperative fluid management often rely on standardized protocols or the surgeon's experience, which may not account for the individual variability among patients.
The objective of this thesis is to create an advanced predictive model using deep learning techniques to forecast the likelihood of anastomotic leakage. The model will take into account various patient-specific parameters and determine the optimal volume of intraoperative fluids to inject, thereby reducing the risk of leakage. The model will be trained on extensive tabular clinical data from multiple hospitals, encompassing preoperative, intraoperative, and postoperative characteristics, annotated by medical experts.
Nature of the Thesis
Programming: 80%, Documentation: 20%
Specific Requirements
• Experience in machine learning
• Good programming skills (especially Python)
Supervisors
Vincent Ochs (PhD student), University of Basel, Center for medical Image Analysis and Navigation (CIAN)
Dr. med. Anas Taha, University of Basel, Center for medical Image Analysis and Navigation (CIAN)
Prof. Dr. Philippe Cattin, University of Basel, Center for medical Image Analysis and Navigation (CIAN)
Contact
Vincent.ochs@unibas.ch
Anas.taha@unibas.ch