Autonomously Adaptive Soundscapes for Reducing Stress in Critically-Ill Patients

• Author(s) / Creator(s)

  • Frishkopf, Michael
  • Papathanassoglou, Elizabeth
  • Hindle, Abram
  • Kutsogiannis, Demetrios

• NFRF Exploration awarded in 2020: High stress levels, delirium, and sleep deprivation are common among critically-ill patients and may compromise recovery and survival as well as increase length and costs of hospital stays. Pharmacologic approaches are the usual mode of treatment for these
  conditions, but with non-negligible expense, limited effectiveness, and potentially harmful side effects. Music and sound therapies are low-cost and non-invasive, without dangerous side effects. Research has shown them to be highly effective if customized to the patient. Yet critically-ill patients cannot be expected to cooperate fully with music therapists, who are also scarce and expensive. We thus propose to design, prototype, and test an autonomous system for generating therapeutic soundscapes adapted to the critically-ill patient, aiming to induce relaxation, improve sleep, and reduce anxiety and delirium. Design of the soundscape space of possibilities will require close collaboration between project researchers in music and medicine and will be shaped by experimental results obtained from a large, diverse group of healthy subjects. A binaural sound generator will create an immersive audio experience auditioned through headphones or speakers. A reinforcement-learning approach will guide the search of the soundscape space based on biosignals: observations of the patient’s current state, including response to the current soundscape. Thus the patient will not need to actively engage with the system but rather will passively provide biofeedback. The explore-exploit tradeoff in the search algorithm must be carefully tailored for critically-ill patients so that the search itself does not cause additional stress. Doing so will require close collaboration between the project’s researchers in computer science, medicine, and music as well as extensive experimentation with healthy subjects. Our innovative synthesis of medical ethnomusicology, music therapy, critical-care medicine, and machine learning aims to produce a cost-effective, adaptive, patient-centered approach to stress reduction in critically-ill patients, one that is non-invasive, non-pharmacologic, and free of harmful side effects, resulting in better quality of life for patients and their families, and more efficient use of hospital resources. Our project also offers learning opportunities for students, and promises to spur future synergistic collaborations between medicine, the arts, and computer science, for the public good.

• Date created

  2019-12-10

• Subjects / Keywords

  • NFRF Exploration
  • New Frontiers in Research Fund Exploration
  • Successful grant
  • Relaxation
  • Reinforcement learning
  • Machine learning
  • Critical care medicine
  • Nursing
  • Medical ethnomusicology
  • ICU delirium
  • Music therapy
  • Stress reduction
  • Critical illness stress
  • Musicology
  • Computer Science
  • Enthnomusicology
  • Soundscapes

• Type of Item

  Research Material

• DOI

  https://doi.org/10.7939/r3-xknb-1r84

• License

  ©️Frishkopf, Michael. All rights reserved other than by permission. This document embargoed to those without UAlberta CCID until 2030.