Predicting PTSD symptom severity and subtype with functional MRI and machine learning

Predicting PTSD symptom severity and subtype with functional MRI and machine learning

  • English language proficiency required
  • Amsterdam

Amsterdam UMC, department of Psychiatry

Post-traumatic stress disorder is a highly debilitating psychiatric disorder which is diagnosed based on the presence of specific symptoms related to exposure to a traumatic and often life-threatening event. However, PTSD is a highly heterogeneous condition that can be diagnosed based on 636,120 different symptom combinations, and thus both biological and symptom subtypes may exist within the clinical population of PTSD that could be uncovered with the use of neuroimaging and predictive models.

In this internship, the student will use brain functional MRI data from 1800 PTSD patients to find replicable patterns of brain/symptom correlations using functional connectivity and machine learning. Given the short internship period, coding experience in Python and/or Matlab and knowledge of AI and data-reduction models (sparse canonical correlation analysis, partial least squares, independent component analysis) is required.

Interested in this internship? Please send an e-mail to l.a.vandemortel@amsterdamumc.nl.

To apply for this job email your details to l.a.vandemortel@amsterdamumc.nl