Approximately 1000 people are diagnosed with cancer in the central nervous system in Norway every year (Norwegian cancer registry). Brain cancer is the deadliest cancer type in people under the age of 40 and one of the most common types of cancer in children. The tumors are heterogeneous and there is great variability in aggressiveness, growth rate and treatment responses. Symptoms and prognosis vary widely and can be difficult to predict even for experienced doctors. To improve the clinical understanding of disease courses and treatment responses The Brain tumor registry and biobank of Mid-Norway was established in 2015 by clinicians and researchers at St. Olavs hospital and NTNU. Based on the data in this registry and data from national and international collaborators, researchers at SINTEF develop computer models to automatically analyze images, predict known prognostic factors and disease course using artificial intelligence and machine learning algorithms.
More specifically, we use deep learning methods to:
(1) Develop tools for quantitative radiological assessment of brain tumors. This includes automatic segmentation of pre-operative, intraoperative, post-operative and follow-up MR/CT/US images to
a. quantify initial tumor localization and volume
b. characterize tumor resection during surgery,
c. quantify residual tumor localization and volume
d. assessment of tumor response to radiation/chemotherapy
e. quantify tumor regrowth.
(2) Combine the use of clinical and radiological data for personalized prediction of clinically important parameters such as survival and post-operative functional level.