Radiomics for improved antidepressant treatment selection
Approved Research ID: 60646
Approval date: July 9th 2020
For treatment of major depression, there are multiple antidepressant drugs available. However, finding the right antidepressant drug for the right patient can take a long time. In fact, after one year, only half of the patients have found an effective drug. In the absence of a better selection method, treating physicians take a trial-and-error approach: start with a specific treatment, take 6-12 weeks to observe efficacy, if symptoms are not efficiently reduced, slowly reduce dosage minimizing withdrawal symptoms, start a new treatment and repeat the cycle.
In this project we want to shorten this trial and error time, by investigating whether a combination of brain MRI characteristics from UK biobank, such as hippocampal volume, combined with advanced image analysis approaches and clinical input can predict which patients will not-respond, or will respond to the antidepressant medications. The outcome of this study (an algorithm) will subsequently be applied to a new study at our home institute, to make sure the algorithm works well. In doing so, this project will have great medical (reduction of side-effects of ineffective medications) as well as economic benefits (shorter treatment durations) for society.