Due to copyright issues, an electronic copy of the thesis must be ordered from the faculty. For the faculty to have time to process the order, the order must be received by the faculty at the latest 2 days before the public defence. Orders received later than 2 days before the defence will not be processed. After the public defence, please address any inquiries regarding the thesis to the candidate.
Trial Lecture – time and place
See Trial Lecture.
Adjudication committee
- First opponent: Professor Rick R. van Rijn, Amsterdam UMC, Netherlands
- Second opponent: Consultant, PhD Nils Thomas Songstad, University Hospital of North Norway
- Third member and chair of the evaluation committee: Professor II Kristin Bjørnland, University of Oslo
Chair of the Defence
Professor II Arne Stray-Pedersen, University of Oslo
Principal Supervisor
Senior Consultant Lil-Sofie Ording Müller, Oslo University Hospital
Summary
Magnetic resonance examination of the whole-body (Whole body MRI) is increasingly used in children and adolescents to assess skeletal involvement in e.g., malignant, or inflammatory disorders. However, MRI images of the pediatric skeleton can be difficult to interpret as normal growth-related changes can be confused with pathology. This may lead to unnecessary follow-up examinations, and, in the worst case, result in wrong diagnosis or maltreatment.
In her thesis, Elisabeth von Brandis and colleagues have examined almost 200 healthy children and adolescents with whole-body MRI. Their aim was to assess the normal appearance of the pediatric skeleton on MRI to be able to better distinguish early pathology from normal growth-related findings.
Further, the research group has trained a machine learning algorithm to delineate and color code various MRI signal patterns to investigate the possibility of using artificial intelligence for automated assessment of the growing skeleton.
Von Brandis’ work reveals that MRI signal interpreted as skeletal pathology in individuals with known or suspected skeletal disease can also appear in healthy children and adolescents. Certain anatomical areas are more frequently involved, and specific signal patterns are more commonly observed than others. This study is unique and provides completely new knowledge that can be useful in interpreting ambiguous skeletal findings, and potentially lead to faster and safer diagnosis of musculoskeletal disorders in young patients. Moreover, it may reduce unnecessary anxiety and stress for patients and parents associated with superfluous follow-up examinations.
The study has also shown that it is practically feasible to develop a machine learning algorithm that can assess different signal patterns on MRI of the growing skeleton. This highlights the possibility of using artificial intelligence for a more efficient, consistent, and accurate interpretation of the pediatric skeleton on MRI in the future.
Additional information
Contact the research support staff.