NIH grant: AI-Driven Detection of Addiction-Like Reward Signatures Through Automated Multi-View Behavioral Analysis (1R41DA064482-01)

retained text

NIH grant 1R41DA064482-01: AI-Driven Detection of Addiction-Like Reward Signatures Through Automated Multi-View Behavioral Analysis
Institute: {'code': 'DA', 'abbreviation': 'NIDA', 'name': 'National Institute on Drug Abuse', 'admin_org_id': '12500072', 'admin_funding_url': ''}. PI: Anthony Edward English. Award: $301390.
Start: 2026-07-01T00:00:00.
Project Summary
The opioid epidemic and substance use disorders devastate millions of lives annually through addiction and
overdose. While animal models are essential for understanding reward processing mechanisms, current
behavioral analysis methods severely limit research progress. Existing approaches rely on basic movement
tracking, subjective human annotation, or complex custom solutions, resulting in inconsistent data across
laboratories. The industry standard still uses simple and outdated tracking technology, preventing researchers
from detecting subtle yet crucial behavioral patterns that could reveal key insights about substance abuse
mechanisms. BioSyft has developed an innovative AI behavioral analysis platform combining a customized multi-
view chamber with optimized computer vision models. This system substantially increases to 29 tracking points
across rodents compared to industry standards, enabling high-accuracy automated behavior classification. Built
on a comprehensive and diverse behavioral dataset, our classification algorithms effectively analyze standard
and nuanced behaviors while maintaining cost-effectiveness and research efficiency. Our focus on AI/ML
technology development for digital health and behavioral phenotyping applications ensures the platform
eliminates manual labeling requirements and standardizes behavioral quantification across laboratories. This
proposal aims to validate and expand automated behavioral analysis through a systematic investigation of
reward-related behaviors during Pavlovian conditioning, combining supervised and unsupervised AI approaches
to detect novel behavioral signatures of addiction. By emphasizing the validation of AI/ML technologies and
investigating the extent of their capabilities, the project will establish standardized metrics for reward-seeking
behaviors while discovering previously undetectable patterns that could advance our understanding of substance
abuse mechanisms.

document fetched 2026-07-12 08:14