# Development of a Machine Learning Algorithm to Detect Opioid Withdrawal and Cravings to Increase Treatment Retention

> **NIH NIH R43** · SPARK BIOMEDICAL INC · 2024 · $318,875

## Abstract

ABSTRACT
The use of opioids in the United States has created an opioid crisis that killed over 80,816 people in 2021. An
estimated 136 people die in the USA each day due to overdose involving an opioid. Discontinuation of opioid
use often results in severe opioid withdrawal symptoms (OWS), creating a barrier to opioid cessation and OUD
treatment retention. Without adequate treatment during opioid detoxification, many patients are unable to
discontinue opioid use and stay in treatment, putting them at higher risk for opioid-related overdose. There is a
need to expand treatment options that can be used safely and effectively in the outpatient opioid detox setting
to increase treatment retention and reduce rates of opioid overdose.
Pharmacotherapies for OWS involve full-agonist treatment with methadone, partial-agonist with buprenorphine,
and alpha(2)-agonist with Lofexidine. However, in 2019, only 1.02M (14%) of the 7.6M Americans with OUD
needing treatment received pharmacotherapy treatment. There are currently FDA-cleared medical devices on
the market designed to reduce OWS, including non-invasive neurostimulation. There are currently no wearable
closed-loop devices on the market designed to monitor and treat OWS and cravings. There is a need to expand
treatment options to effective devices that provide tailored, targeted therapy by both monitoring and treating
patients with OWS to increase treatment retention and prevent relapse.
We propose the combination of physiologic sensing and neurostimulation to deliver tailored, closed-loop opioid
withdrawal treatment. Spark Biomedical has developed a wearable transcutaneous auricular neurostimulation
(tANTM) system, Sparrow Ascent. Sparrow Ascent is intended to transcutaneously stimulate branches of the
vagus and trigeminal nerves on and/or around the ear. Sparrow Ascent is based on the Sparrow Therapy
System, which was FDA-cleared in 2021 (K201873) and indicated for use as a transcutaneous nerve stimulator
that aids in the reduction or prevention of OWS. A randomized-controlled trial shows Sparrow can reduce OWS
by 46% after 60-min during acute detox. Spark partnered with Battelle Memorial Institute to integrate a closed-
loop opioid withdrawal and cravings detection algorithm with the NeuroHub platform. NeuroHub couples
wearable sensors to measure heart rate, heart rate variability, cortisol levels, accelerometry, temperature and
electrodermal activity. NeuroHub uses a cloud-based application programming interface (API) and machine
learning (ML) to create the decision tree for automated detection of OWS and cravings based on sensor feedback
to selectively trigger tAN treatment in a closed-loop system.
In this Phase I, we propose to develop an opioid withdrawal detection and cravings algorithm using
wearable physiologic sensors. The long-term goal for this algorithm is to enable patient monitoring and inform
the closed-loop delivery of tAN therapy to reduce OWS and cravings to increase treatment retent...

## Key facts

- **NIH application ID:** 10783563
- **Project number:** 1R43DA059448-01
- **Recipient organization:** SPARK BIOMEDICAL INC
- **Principal Investigator:** Navid Khodaparast
- **Activity code:** R43 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $318,875
- **Award type:** 1
- **Project period:** 2024-06-15 → 2026-05-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10783563

## Citation

> US National Institutes of Health, RePORTER application 10783563, Development of a Machine Learning Algorithm to Detect Opioid Withdrawal and Cravings to Increase Treatment Retention (1R43DA059448-01). Retrieved via AI Analytics 2026-05-24 from https://api.ai-analytics.org/grant/nih/10783563. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*