Cognitive-behavioral therapy (CBT) is the first-line treatment for obsessive-compulsive disorder (OCD), yet a significant number of patients do not achieve remission or substantial symptom relief. This study aims to enhance the prediction of CBT outcomes in OCD by integrating demographic, clinical, and neuroimaging data using machine learning (ML) models. We conduct a comprehensive analysis on a well-characterized clinical sample, employing a rigorous validation scheme to avoid data leakage, and comparing multiple ML algorithms to minimize bias. Out of four different ML models trained on demographic and clinical data, structural MRI, and resting-state MRI functional connectivity data, no model was able to predict CBT success significantly above chance level in the present sample. Although clinical and demographic data enabled 64%-66% accuracy for predicting remission, this did not reach statistical significance after permutation testing. Pre-treatment symptom severity emerged numerically as the most promising predictor of remission, aligning with previous studies, but did not pass the significance threshold in the present study. Despite efforts to identify neuroimaging predictors, neither functional nor structural MRI features significantly contributed to the prediction models. These findings suggest that robust, individualized brain-based predictions for mental health outcomes remain challenging with the available data and sample size.
Marija Tochadse, Julia Klawohn, Christian Kaufmann et al.· Scientific Reports· 0 citations
Control variables are widely used in statistical modelling to account for omitted variable bias of known confounders. However, they have largely been underexplored in deep learning. This is surprising, given that deep learning models encode image-inferable covariates, such as demographic variables, into their predictions when these covariates are correlated with the outcome---a form of omitted variable bias referred to as'shortcut learning'. While many existing confound-control or fairness methods try to restrict the correlation of such covariates with model predictions, we show that this fails to correct for omitted variable bias. We therefore propose a control variable approach for deep learning models, based on generalised additive modelling of the effects of model inputs and covariates. As flexible additive models can suffer from concurvity, we introduce an estimation procedure that refits the final layer of a pre-trained network to include covariate effects, using cross-fitting with ridge penalisation. We show how these effects can be orthogonalised with respect to covariates to exclude their mediated effects and that model predictions can be marginalised over the covariate distribution to control for their effect. This yields unbiased, interpretable predictions and offers flexibility to model the desired effects depending on the scientific or fairness objective. We verify our approach using simulated images, and demonstrate consistent estimation of true effects. Existing methods either require more data or fail to recover the true effects. We apply our method to real neuroimaging data with experimentally induced confounding, where it recovers prediction performance to near the level of a model trained on unconfounded data. Code is available at https://github.com/mpff/cocodeel.
Manuel Pfeuffer, R. Rane, Kerstin Ritter et al.· 0 citations
CON decomposition is introduced, which quantifies how much of a layer's variance each concept explains given all other concepts and the outcome, and how much none of them explains, yielding layer-comparable, calibrated scores that suppress false positives.
R. Rane, Marco Simnacher, Manuel Pfeuffer et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
