Machine Learning Approaches for Clinical Psychology and Psychiatry

Abstract

Machine learning approaches for clinical psychology and psychiatry explicitly focus on learning statistical functions from multidimensional data sets to make generalizable predictions about individuals. The goal of this review is to provide an accessible understanding of why this approach is important for future practice given its potential to augment decisions associated with the diagnosis, prognosis, and treatment of people suffering from mental illness using clinical and biological data. To this end, the limitations of current statistical paradigms in mental health research are critiqued, and an introduction is provided to critical machine learning methods used in clinical studies. A selective literature review is then presented aiming to reinforce the usefulness of machine learning methods and provide evidence of their potential. In the context of promising initial results, the current limitations of machine learning approaches are addressed, and considerations for future clinical translation are outlined.

Brain-predicted age difference score is related to specific cognitive functions: A multi-site replication analysis

“…For practical reasons, crossvalidation, where a dataset is split into a training set and a test set , is often used as an estimate of model accuracy for new data (Jollans & Whelan, 2018;Scheinost et al, 2019). However, cross-validation accuracy estimates are often optimistically biased and can vary considerably , particularly when preprocessing and feature selection are carried out on the entire dataset before splitting it into training and test sets (Dwyer, Falkai, & Koutsouleris, 2018;Woo et al, 2017). As such, the gold-standard for assessing the external validity and generalisability of a neuromarker is by testing how the model performs on a completely independent held-out dataset (Jollans & Whelan, 2018).…”

Scaling up psychology via Scientific Regret Minimization

“…On a broader note, we hope to further the development of a synergistic correspondence between psychology and data science approaches in scientific modeling (Rosenfeld, Zuckerman, Azaria, & Kraus, 2012;Dwyer, Falkai, & Koutsouleris, 2018;Bourgin, Peterson, Reichman, Griffiths, & Russell, 2019). Cognitive science famously grew out of the intersection of six different fields (Gardner, 1987), but some have suggested that this revolution did not create the SCIENTIFIC REGRET MINIMIZATION 26 emergence of a new discipline (Lakatos, 1986;Miller, 2003;Núñez et al, 2019).…”

Precision non-implantable neuromodulation therapies: a perspective for the depressed brain

“…These approaches are important and fundamentally different from statistical approaches to identification of predictors, which are based on groups, not individuals. 13,147 In fact, statistical methods focus on inference -creating a mathematical model that tests a hypothesis about how a system behaves, whereas machine learning focuses on prediction -i.e., finding generalizable predictive patterns that aim to forecast future behaviors regardless of their mechanistic basis 12 (Figure 3). Additionally, through employing almost no pre-assumptions and a nonlinear function canvas, machine learning techniques can model complex patterns that can identify relationships between large amounts of data and data of diverse types, 148,149 increasing the processing speed and output of predictive models.…”

“…6 Although this concept is not necessarily new (e.g., blood transfusion is ''guided'' by blood type examination), three new emerging tools [6][7][8] are involved in the precision psychiatry framework: 1) incorporating the biological pathways of disease -in psychiatry, this is represented by the National Institute of Mental Health (NIMH) Research Domain Criteria (RDoc), a framework that evaluates mental illness at multiple clinical, endophenotypic, and neurobiological levels 9 ; 2) multimodal big data collection, i.e., acquisition of clinical and biological data at scale, as exemplified by the opportunities presented by international consortiums such as the Enhancing NeuroImaging Genetics through Meta-Analysis (ENIGMA) 10 and megacohorts such as the UK Biobank 11 ; and 3) artificial intelligence for analysis of multidimensional and complex patterns in manifold data collected at multiple biological levels. 12,13 Although precision psychiatry is still in its infancy, the continuous, rapid development of these tools will reshape clinical and research practice, enhancing treatment and minimizing adverse effects. 6 Non-implantable neuromodulation (NIN) interventions, such as transcranial magnetic stimulation (TMS), transcranial direct current stimulation (tDCS), electroconvulsive therapy (ECT), and magnetic seizure therapy (MST), are non-pharmacological, non-psychotherapeutic interventions with distinct efficacy, safety, tolerability, and availability profiles.…”