SINAI-PD

Synchronous Intraoperative AI-based decision-making in partial duodenoapancreatectomy

Removal of the pancreatic head for cancer or chronic inflammation is one of the riskiest visceral surgical operations with a postoperative mortality rate of approx. 5-8% in specialized centers. In the standard operation, partial duodenopancreatectomy, the tail of the pancreas is preserved and connected to the small intestine. This prevents diabetes, but carries a high risk of pancreatic fistulas (30%), which can lead to life-threatening complications. A decision must be made intraoperatively as to whether it makes sense to preserve the pancreatic remnant (with possible complications) or whether complete removal (with subsequent diabetes) is necessary. This decision is based on the intraoperative assessment of tissue quality and has far-reaching consequences for the postoperative outcome of the patient. The overarching goal of the project is to develop a decision support system using artificial intelligence (AI), that combines intraoperative image analyses and perioperative patient data to enable an objective risk assessment. This can minimize subjective influences and sustainably improve patient safety through data-based decision-making.

The aim of the project is to establish interdisciplinary cooperation between the Department of Surgery at Giessen University Hospital (Prof. Hecker, Dr. Anca) and the KITE Competence Center (Prof. Guckert, Prof. Hannig), which will form the basis for further applications for third-party funding initiated by this project. The project pursues an interdisciplinary approach in which state-of-the-art image processing technologies are combined with clinical expertise in order to establish evidence-based, AI-supported support for surgical decision-making situations.

  Prof. Andreas Hecker, Justus Liebig University Giessen, University Hospital Giessen & Marburg GmbH

Prof. Michael Guckert, Technische Hochschule Mittelhessen, KITE Dr. Anca-Laura Amati, University Hospital Giessen & Marburg GmbH Dr. Jennifer Hannig, Technische Hochschule Mittelhessen, KITE, Junior Research Group TimeXAI