Work

Predicting treatment response in depression: insights from electroencephalography studies

Type: article
Abstract: Major depressive disorder (MDD) is one of the most prevalent mental illnesses, significantly affecting individuals and society.Despite the availability of effective first-line treatments, such as pharmacological and psychotherapeutic options, there remains a lack of reliable methods to predict which treatment will benefit individual patients.This forces practitioners to rely on a suboptimal trial-and-error approach, resulting in low response rates following initial treatment attempts.Additionally, it can take up to twelve weeks for treatments to show effects, which further delays recovery.These challenges highlight the urgent need for innovative strategies to improve treatment outcomes and alleviate the burden of MDD.Recently, neuroimaging research has played a pivotal role in identifying neural biomarkers that can predict treatment response.Among these, electroencephalography (EEG) stands out as a non-invasive technology that provides unique insights into brain activity patterns with millisecond resolution.While several studies have attempted to develop predictive models utilizing EEG biomarkers, their clinical validity is yet to be established.Both advanced neuroengineering solutions and further research are required for successfully integrating these models into clinical practice and demonstrating how EEG technology can support personalized treatment approaches, which together form the core of this thesis.We first focused on validating EEG-based predictive models and facilitating their translation to clinical settings.To this end, we developed the first machine learning model that demonstrated generalizability in predicting response to medications across two large independent EEG datasets.Following that, we established standardized EEG methodologies across multiple clinical sites, providing a framework essential for the real-world evaluation of biomarkerinformed predictive models.We then illustrated that the clinical utility of EEG encompasses more than just predicting response to medications.Specifically, we identified EEG patterns differentiating responders from non-responders to cognitivebehavioral therapy, offering the first evidence of EEG biomarkers for this commonly prescribed treatment.We also examined changes in EEG occurring during first-line interventions, revealing unique neural mechanistic biomarkers related to symptom improvements with medication treatments that could serve as potential targets for more personalized therapies.Collectively, these four studies significantly enhance MDD care by bridging the gap between EEG research and its real-world clinical application.
Language: English
Field: Medicine
Open Access status: green
OpenAlex ID: W7139826597
Cites: 302
Cited by: 1
Related to: 10