Lisanne van Dijk

UMC Groningen

In short, the research of Lisanne van Dijk entails the following:

AI-based lymph node-specific risk stratification: A clinical decision tool for reducing post-chemoradiotherapy salvage complications in head and neck cancer

In 20–40% of patients with head and neck cancer with regional lymph node metastases treated with curative (chemo)radiotherapy, post-treatment imaging raises suspicion of residual nodal disease, leading to salvage neck dissection. However, in 60–70% of these cases, pathology reveals no residual tumor. Salvage surgery after radiotherapy is extensive and associated with a high complication risk, resulting in substantial and often unnecessary morbidity. Hence, this post-radiotherapy surgery should be avoided whenever possible.

This grant aims to develop AI models that can predict, already before treatment, which individual lymph nodes are at high risk of having residual tumor after radiotherapy. Radiomics and deep learning from CT, PET, and MRI will be combined to enrich these models. Subsequently, a clinically usable decision-support tool will be developed to estimate the risk per lymph node, enabling more precise treatment of high-risk nodes and helping to avoid unnecessary salvage neck surgery.