Tomas Janssen

NKI

In short, the research of Tomas Janssen entails the following:

Identification of inaccurate retroperitoneal sarcoma delineations to reduce the risk of local failure

Retroperitoneal sarcoma (RPS) is a rare cancer in which accurate definition of the tumor is essential for effective radiotherapy. In fact, studies have shown that when the tumor is drawn inaccurately, patients face a substantially higher risk of a local recurrence, sometimes with fatal consequences. Ensuring consistent and accurate delineation is challenging, even for experts. Moreover, no practical quality-assurance tools currently exist in routine care. This project addresses that gap by developing artificial-intelligence (AI) technology that can automatically detect when a tumor delineation is likely to be incorrect, helping clinicians avoid potentially harmful mistakes.

Within the project, we will train and validate an AI-based quality-assurance system using data from the STRASS trial and several international cancer centers. Once validated, the tool will be implemented in the Netherlands Cancer Institute to support clinicians during daily decision-making. By flagging high-risk cases that need extra review, the system aims to make radiotherapy for RPS safer, more consistent, and more effective. Ultimately, this project will help reduce the chance of local cancer recurrence and improve outcomes for patients facing this challenging disease.