The CLARITY project (Composite Longitudinal Assessment of tumor Response Integrating Tumor volume and segmentation uncertainty) aims to develop and validate a novel AI-driven imaging biomarker framework that goes beyond traditional tumor size measurements. Instead of relying solely on tumor diameter (RECIST 1.1), the project leverages deep learning to quantify complex 3D disease characteristics, internal tumor structures and “boundary clarity”.
The project focuses on two important case studies:
Pancreatic Ductal Adenocarcinoma (PDAC) and
Colorectal Cancer Liver Metastases (CRLM).
CLARITY is a highly collaborative two-year project led by
Dr. Aneta Lisowska. You will work in a multidisciplinary research environment at
Amsterdam UMC, bridging the fields of Surgery, Radiology and Radiation Therapy. The team brings together clinical and technical experts to ensure that the developed AI tools are clinically relevant and aligned with real-world medical decision-making.
As a researcher within this project, your main goal is to translate complex AI challenges into clinically meaningful tools that improve cancer treatment monitoring. You will contribute to the development of the
Composite Response Index (CRI), a novel biomarker that integrates traditional tumor volume changes with a
Margin Clarity Score (MCS) derived from AI-based uncertainty models.
Your key responsibilities include:
- Correlating clinical ambiguity in tumor delineations with patient treatment responses using radiotherapy and surgical datasets.
- Benchmarking different AI uncertainty methods to determine the most effective approach for capturing clinical disagreement.
- Developing and validating the CRI biomarker using large-scale clinical trial cohorts.
- Building interpretable predictive models using Explainable AI (XAI) to support clinical decision-making.
- Collaborating closely with clinical end-users in dedicated sessions to refine the framework and ensure it fits multidisciplinary team workflows.