Browsing by Author "van der Wijk, Gwen Marijke"

Now showing 1 - 1 of 1
  • Thumbnail Image
    ItemOpen Access
    Considering Individual Variation in the Search for Neuroimaging Features of Response to Pharmacotherapy for Major Depression
    (2022-09) van der Wijk, Gwen Marijke; Protzner, Andrea B.; Kiss, Zelma; Dobson, Keith Stephen; Gratton, Caterina; Ismail, Zahinoor
    Many studies investigate the potential of neuroimaging features to predict treatment success for patients with major depression (MD), as this could improve outcomes in clinical practice. However, variable findings and methodological issues limit the generalizability of such research. In addition, we know little about the applicability of features identified in neuroimaging studies at an individual level, which is essential for translation to clinical practice. In this thesis, I examined these issues to advance our understanding of the opportunities and challenges that this field can leverage to get closer to personalized care. First, I investigated the robustness of fMRI functional connectivity features in three important brain networks related to MD and successful treatment using a large, multi-site dataset. I identified stable differences between participants before treatment based on if and how quickly their symptoms decreased during treatment, indicating their potential to predict outcomes. Second, I explored the relative magnitude of individual variation and group differences such as those identified in project 1. Specifically, the similarity in whole-brain fMRI connectivity across everyone, groups (patients vs controls, responders vs non-responders, female vs male participants), sessions (baseline, week 2 and 8) and individuals was quantified to estimate the relative amount of variance explained by each of these sources. Individual-specific connectivity, together with common connectivity across participants and sessions, explained most of the variance in the data, while group differences contributed only a small amount. Third, I examined the group-to-individual generalizability of brain features using EEG. After identifying differences between groups of patients whose symptoms did or did not decrease substantially with treatment, this study explored whether such group features could be identified in individual patients. The results revealed that individual brain features often deviated from group features. Overall, these findings indicate that, though robust features of antidepressant treatment success may be identified at the group level using large samples and thorough standardizing procedures, individual variation likely needs to be considered for these findings to be applicable to individual patients. Future research should examine if individual brain features can accurately inform clinical practice.