Surgery is a central pillar of modern oncology. At the same time, its success strongly depends on the performing surgical team’s level of experience. Research suggests that surgeries performed by less experienced staff exhibit higher rates of complication, a large proportion of which could, in principle, be avoided. 

The driving hypothesis of our research at the Surgical AI Division is that artificial intelligence (AI) based on surgical data science has the potential to revolutionize surgery by systematically elevating its safety, efficiency, and quality. Committed to the ultimate goal of creating benefit for both patients and clinical staff, and building upon principles and knowledge from diverse research fields including machine learning, statistics, computer vision, biophotonics, and medicine, we seek to harness the power of data within and beyond the operating room to develop intelligent systems for interventional cancer care. Our research particularly focuses on the intraoperative setting, a challenging environment in which clinical staff need to take in a multitude of different information from various sensors and devices while navigating complex anatomical structures, often under conditions of restricted view, make clinical decisions, and react to unforeseen events at any moment. Our research aims to facilitate this process and thereby enhance both surgical performance as well as safety. In terms of applications, we are specifically using the power of deep learning to develop modern imaging concepts based on biophotonics techniques that allow the non-invasive measurement of functional tissue parameters such as blood oxygenation deep within tissue. Another focus lies on the development of decision support systems that are able to integrate multimodal sensor and other patient data with factual and experiential knowledge to assist surgeons in intraoperative decision making in real time.