Psysensor: Integrating Biopsychosocial Profiles for Treatment of Mood and Anxiety Disorders

Overview

The Diagnostic and Statistical Manual of Mental Disorders (DSM) has long been the cornerstone of psychiatric diagnosis, yet its relevance in the modern era of 2024 is increasingly questioned. The initial challenge arises as patients enter a clinic, their symptoms often so ambiguous that agreement between clinicians on the same patient's symptoms is remarkably low. The DSM's reliance on superficial symptom-based labels further complicates the issue, leading to disparate diagnoses for patients with similar underlying issues and identical diagnoses for fundamentally different patients based on surface similarities. This turns the selection of appropriate treatment into a gamble, leaving countless patients at a disadvantage, with studies showing the DSM has as low as a 46% accuracy or 64% with a psychiatrist's input and over 80% of patients waiting months, years, or decades to receive treatment. This study introduces and evaluates a novel method for mental health care, moving beyond the DSM's subjective symptomatology to an objective, data-driven model. N=500 patients underwent brain MRI scans, participated in seven psychological evaluations, and completed three social assessments, creating a comprehensive dataset of 128 biopsychosocial variables per patient, or a unique "fingerprint." Six artificial intelligence models analyzed these fingerprints to find patterns correlating with successful treatment outcomes, bypassing the traditional trial-and-error approach. The structure of the most efficient model was implemented in a new algorithm which predicted the most effective treatment with 93.8% accuracy, significantly surpassing the DSM's performance by 56% (p<.001). This rapid and precise treatment allocation in seconds rather than years promises to eliminate the harmful consequences of the DSM's inefficiencies, including unnecessary deaths, treatment delays, and financial burdens, marking a pivotal shift towards a more effective, data-driven future in mental health care.