Collaborative Research: SCH: Psychophysiological Sensing to Enhance Mindfulness-Based Interventions for Self-Regulation of Opioid Cravings

Opioid use disorder (OUD) is a chronic condition and a leading public health problem in the U.S. The risk of overdose is particularly high following a period of abstinence leading to drug-related deaths. OUD includes physical dependency and neural adaptations in brain circuits of reward and motivation, self-regulation, and stress reactivity that can persist years after drug discontinuation.

Substance craving is one of the primary causes of OUD patient's relapse. Studies have shown psychological cues such as stress, anxiety, and arousal can precipitate the cultivation of drug craving. Research has found that mindfulness-based strategies reduce cravings, psychological cues and prevent relapse.

Mindfulness-based interventions (MBIs) bring about clinically relevant changes to physiological arousal, stress, and addictive behavior through cognitive behavioral skill development. This project focuses on developing and testing innovative technologies to aid sustainable recovery of OUD with wearable and in-home physiological monitoring and generation of adaptive, personalized, and just-in-time MBIs.

While the research is focused on OUD, the principle and the outcomes can be expanded to include other substance use disorders. The project includes several education and outreach activities such as machine learning course for medical professionals and annual workshops for middle school girls.

This study focuses on opioid use disorder (OUD), related cognition, and behaviors associated with a) reward, b) self-regulation, c) stress reactivity, d) opioid craving, e) physical opioid withdrawal symptoms and MBIs known to be impacted by OUD and post-acute withdrawal from opioids. In particular, the research tasks focus on there areas. First, effective physiological feature identification and extraction to detect craving that is generalizable across large OUD populations and consider the external factors such as age, gender, drug use habits, etc.

Second, development of an effective multi-modal sensing integration approach to capture psychological craving cues (e.g., stress, arousal) from a combination of acoustic and physiological sensing. This will include novel multiple instance (MIL) multitask learning based classification techniques that are scalable with near real-time performance. The study will address the fundamental gaps of indoor craving-relevant sensing where only a small fraction of a long signal may convey information relevant to the targeted emotional state/class.

The last task will include development of a craving context-aware MBI recommender system that models the dynamic nature of OUD subjects craving-interventions and feedbacks. The system will be formally validated to ensure safety against adverse outcomes. Successful execution of the research will begin to test the effectiveness of integrating passive sensing, adaptive artificial intelligence (AI), and mindfulness interventions on regulating drug craving.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.