Optimizing DIEP flap outcomes using thermographic imaging
Breast reconstruction using autologous tissue, specifically with Deep Inferior Epigastric artery Perforator (DIEP) flaps, has made significant progress in recent years by minimizing donor-site damage. Selecting the appropriate perforator is crucial, as the flap is supplied by only one perforator. Although Computed Tomography Angiography (CTA) is the standard method for vessel selection, it has several drawbacks, including the use of contrast agents, exposure to radiation, high costs, and the lack of information about flap perfusion. Recent studies have shown that Dynamic Infrared Thermography (DIRT) is a non-invasive technique that can visualize dominant perforators preoperatively and perfused zones intraoperatively. Identifying these zones is essential to optimize flap survival, but this currently requires an additional average surgical time of around 60 minutes.
To address this challenge, the project explores two complementary paths for improvement. The first focuses on detecting the most optimal perforator choice preoperatively using Dynamic Infrared Thermography (DIRT) combined with artificial intelligence. The second represents a groundbreaking approach: eliminating the need for intraoperative measurements entirely by predicting perfused zones of specific perforators before surgery. This innovative solution leverages a Convolutional Neural Network trained on advanced Finite Element Method (FEM) models of the abdomen with perforators. These models integrate CTA data with pre- and intraoperative DIRT measurements, and employ FEM updating to accurately replicate thermal behavior observed through infrared imaging. By shifting from time-consuming intraoperative assessments to predictive modeling, this approach has the potential to redefine surgical planning and significantly improve efficiency and patient outcomes.
