Explainable AI-driven precision clinical trial enrichment: demonstration of the NetraAI platform with a phase II depression trial.
Joseph Geraci, Bessi Qorri, Mike Tsay, Christian Cumbaa, Paul Leonczyk, Larry Alphs, Elizabeth D. Ballard, Carlos A. Zarate, Luca Pani
NPJ Digital Medicine December 8, 2025 DOI: 10.1038/s41746-025-02143-7 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Secondary analysis of a Phase II clinical trial Peer reviewed |
|---|---|
| Sample size | 63 |
| Population | Patients with treatment-resistant depression enrolled in a Phase II ketamine trial |
| Intervention | Ketamine |
| Topics | Depression |
| Key findings | NetraAI, an explainable AI platform, predicted ketamine treatment outcomes in treatment-resistant depression better than standard machine learning, improving accuracy by approximately 25-30% and identifying a 10-clinical-variable model with AUC 0.32 higher than standard ML and an 8-MRI-feature model with 95% accuracy and 100% specificity. The authors argue this demonstrates that dynamical AI can uncover clinically meaningful patient subgroups from small, rich datasets. |
Abstract
Clinical trial failures are frequently driven by patient heterogeneity and limited sample sizes that obscure treatment effects by diluting statistical power. We introduce NetraAI, a novel explainable artificial intelligence (AI) platform that integrates dynamical-systems modeling, evolutionary long-range memory feature selection, and large-language model (LLM)-generated insights, to discover high-effect-size patient subpopulations ("Personas") from high-dimensional clinical data. In a Phase II ketamine trial for treatment-resistant depression (n = 63), NetraAI analyzed psychiatric scale data (175/patient) and MRI-derived features (185/patient). NetraAI outperformed traditional machine learning (ML) models in predicting treatment outcomes, improving predictive accuracy by approximately 25-30% and achieving higher sensitivity and specificity in detecting responders. NetraAI identified a 10-clinical variable model that improved predictive AUC by 0.32 over standard machine learning (ML) models and an 8-MRI feature model achieving 95% accuracy and 100% specificity. These findings demonstrate that an explainable dynamical AI approach can leverage small but rich datasets to uncover hidden clinically meaningful subgroups. NetraAI's precision enrichment strategy has the potential to improve trial success rates and enable personalized medicine by prospectively identifying patients most likely to benefit from a given therapy in oncology, psychiatry, neurodegeneration, and for other disorders.
Comparable studies
Other non-randomized and open-label trials on depression, most cited first.
Citations in the library
Cites 6
- Measurement of dissociative states with the Clinician‐Administered Dissociative States Scale (CADSS) Journal of Traumatic Stress 1998
- Ketamine and Rapid-Acting Antidepressants: A Window into a New Neurobiology for Mood Disorder Therapeutics Annual Review of Medicine October 23, 2014
- Ketamine Has Distinct Electrophysiological and Behavioral Effects in Depressed and Healthy Subjects Molecular Psychiatry February 27, 2018
- International pooled patient-level meta-analysis of ketamine infusion for depression: In search of clinical moderators Molecular Psychiatry September 7, 2022
- Brain-based correlates of antidepressant response to ketamine: a comprehensive systematic review of neuroimaging studies. The Lancet Psychiatry October 1, 2023
and 1 more in the library