NetraMark Presents AI-Driven Advances in Precision Psychiatry to Enhance Clinical Trial Designs at Joint Autumn Conference

Executive Summary

• NetraMark presented two AI‑driven applications at major CNS clinical‑trial conferences, demonstrating how its NetraAI platform can differentiate true drug effects from placebo and improve predictive accuracy in MDD trials.
• In a ketamine Phase II trial, NetraAI identified distinct responder subgroups with increasing separation (0.75 → 1.0) and cohesion (~25% → ~67%) after the second infusion, supporting genuine pharmacologic signal.
• Using data from the CAN‑BIND escitalopram study, NetraAI reduced variables to eight key features, uncovered a hypomethylated responder subgroup, and achieved 91% prediction accuracy for treatment response.

Key Details

• Ketamine Trial Findings
• Cross‑group divergence rose from 0.75 (moderate) after infusion 1 to 1.0 (complete separation) after infusion 2.
• Within‑group cohesion for ketamine responders increased from 25% → ~67% across the two infusions.

• Demonstrates that responder subtypes drive efficacy, not functional unblinding.

• Escitalopram (CAN‑BIND) Trial Findings

• Traditional ML used 718 clinical variables with modest accuracy; NetraAI distilled to 8 key variables (anhedonia, daily functioning, appetite, negative thinking).
• Classification produced 26 non‑responders, 23 responders, and 124 “unknown” cases.
• Methylation analysis on 169 patients (≈68,600 CpG sites each) identified 33 of 117 responders as “Highly Predictive Responders” (HPRs).

• Hypomethylation at three gene loci (LINC01580, STK24, ATXN7L3) linked to neuroplasticity; models retrained on these features predicted response with 91% accuracy.

• Strategic Implications

• NetraAI’s explainable ML can enable persona‑guided enrichment, stratified randomization, and more efficient CNS trial designs.
• Potential to reduce sample sizes, accelerate timelines, and improve drug‑development success rates in psychiatric indications.

Notable Quotes

• “The ability to distinguish true pharmacologic effects from placebo and to predict treatment response with explainable subgroups represents a meaningful advancement,” said Josh Spiegel, President of NetraMark.