Wearable Automatic Pain Assessment For Olfaction-Based Chronic Pain Management
2024-03
Loading...
View/Download File
Persistent link to this item
Statistics
View StatisticsJournal Title
Journal ISSN
Volume Title
Title
Wearable Automatic Pain Assessment For Olfaction-Based Chronic Pain Management
Authors
Published Date
2024-03
Publisher
Type
Thesis or Dissertation
Abstract
More than 67 million Americans -- about 20% of the U.S. population -- suffer from chronic pain. In addition to the suffering that it induces, chronic pain also leads to higher prevalence of depression, insomnia, and physical and mental impairment. A study showed the estimated cost for chronic pain management was between $560-635 billion in 2010 dollars. Despite the significant problems that chronic pain presents, there is a lack of effective pain management solutions. Medication-based treatments are not completely effective, can lead to drug overdoses, and play a major role in the current opioid crisis. Mind and body interventions (MBIs) -- activities that change mental and physical state -- have been shown to be an effective pain management solution without the negative side effects associated with typical approaches. One particular MBI is using a scent to help patients develop and maintain a mental association between the scent and a lower pain state that can later be used to manage or prevent a pain episode. This approach relies on an automated pain assessment and scent delivery device that can accompany a patient wherever they go. The most logical implementation involves a wearable device to monitor physiological signals related to pain and dispense scents as necessary. In this work, we describe a wearable device that monitors various physiological signals and processes them to predict a user's pain state. Various machine-learning models are trained with the data collected from clinical trials using a prototype of the wearable device. The correlation between physiological signals and pain agrees with results from literature, but the performance of machine learning models is not ideal due to limited dataset size. Results suggest that collection of additional physiological data for model training will increase performance. This work represents an important step toward novel, safe, and effective MBI-based chronic pain management.
Keywords
Description
University of Minnesota M.S. thesis. March 2024. Major: Electrical/Computer Engineering. Advisor: John Sartori. 1 computer file (PDF); xi, 52 pages.
Related to
Replaces
License
Series/Report Number
Funding information
Isbn identifier
Doi identifier
Previously Published Citation
Suggested citation
Content distributed via the University Digital Conservancy may be subject to additional license and use restrictions applied by the depositor. By using these files, users agree to the Terms of Use. Materials in the UDC may contain content that is disturbing and/or harmful. For more information, please see our statement on harmful content in digital repositories.