Projects

Projects

 

 

Research Agenda: AI-Driven Surveillance of the Opioid and Fentanyl Epidemic 

Principal Investigator: Dr. Yulin Hswen
Research Focus: Leveraging AI and social media data to monitor and predict the dynamics of the fentanyl epidemic and illicit drug trafficking

1. Introduction

The fentanyl epidemic continues to evolve, requiring innovative surveillance methods to understand its spread and impact. Traditional epidemiological models rely on delayed reporting, whereas real-time social media and digital data can provide timely insights. This research aims to harness AI-driven methods to analyze digital discourse, identify emerging drug trends, and assess the effectiveness of public health interventions without reinforcing stigma against affected populations.

2. Key Data Sources

  • Social Media Platforms (Meta, TikTok, Reddit, X): AI-driven sentiment and network analysis to track discussions on fentanyl use, emerging analogs, and harm reduction strategies.
  • The Street Drug Analysis Lab at UNC: Data integration to map real-world drug composition trends against social media signals.
  • Online Marketplaces and Dark Web Monitoring: NLP models to track illicit drug supply chain movements.
  • Law Enforcement and Public Health Reports: Comparative validation of AI-generated trends with official seizure and mortality data.
  • Counterfit and Drug Diversion of Perscription Opioids: Using data crawling and social sentinel information to track fake and cutting of drugs into pure sources.

3. AI Methodologies for Surveillance

  • Natural Language Processing (NLP): Analyzing online discourse for shifts in fentanyl discussions, new synthetic opioids, and harm reduction awareness.
  • Computer Vision and Image Recognition: Detecting visual cues related to drug paraphernalia and coded language in digital spaces.
  • Social Network Analysis (SNA): Identifying influential nodes in online drug markets and the spread of misinformation.
  • Geospatial Mapping: AI-driven predictive models to map fentanyl hotspots and inform targeted intervention strategies.
  • Multimodal Integration: Combining textual, image, and network data to refine surveillance accuracy.

4. Policy Relevance and Applications

This research informs policy in several critical areas:

  • Public Health Messaging: AI insights can refine messaging strategies to enhance engagement.
  • Predictive Risk Assessment: Policymakers can use AI-driven early warnings to implement timely interventions.
  • Supply Chain Disruption: Understanding digital traces of fentanyl trafficking can inform international policy collaborations to disrupt supply chains.

5. Conclusion

Dr. Yulin Hswen's AI-driven surveillance framework provides a critical, lens on the fentanyl crisis, leveraging social media and digital data to inform real-time interventions and evidence-based policymaking. By integrating computational epidemiology with digital public health strategies, this research enhances societal resilience against the evolving drug crisis.