︎
The University of Michigan
+ The Mayo Clinic
Sonification for Sleep - Sonic PhenotypingUARTS Faculty Engineering/Arts Student Team (FEAST)
sleep ︎ data ︎ music
︎
ABOUT
Working with doctors at the Mayo Clinic Center for Sleep Medicine, our UARTS Faculty Engineering/Arts Student Team (FEAST) is exploring the possibilities of creating techno tracks from up to, at least, fourdata points from raw polysomnogram data (EEG/Pulse/Oxygenation). The goal is to convert sleep data into interesting music to enable sleep diagnostics that would be accurate and fun for the world. Sonification data has already been shown to increase the speed and accuracy of interpretation of medical data and we aim to expand that principle to the world of sleep diagnostics.
Introductory experiments, working with CoE alumnus Greg Syrjala (from ORMEC in Rochester, NY) have been conducted, in collaboration with Dr. Kara DuPuy-McCauleyat the Mayo Clinic. Kara is an alumnus of the School of Music, Theatre & Dance, and had an established career in Punk band before returning to Michigan to complete her studies in Medicine. It is her life’s vision to create an “audio approach” for sleep study data and using MAX/MSP, our team has already concatenated EEG, EKG, EMG and RespNatal data using EDF formatted files introduced into MAX/MSP and Ableton live. The results sounded like “glitch” meets “Brian Eno/ambient music.” The goal is to create interested music with data that is inherently rather dull and does not change much, but highlight anomalies, such as snoring (or worse), in order to diagnose sleep patients more accurately and efficiently. Sonification of data has been utilized for decades in the medical field but although we can hear the representation of patient data rather ubiquitously in the operating room or intensive care unit (think beeps from EKG and ventilators), sonification as a field of medical research is relatively unexplored in most other specialties.
zzz
Introductory experiments, working with CoE alumnus Greg Syrjala (from ORMEC in Rochester, NY) have been conducted, in collaboration with Dr. Kara DuPuy-McCauleyat the Mayo Clinic. Kara is an alumnus of the School of Music, Theatre & Dance, and had an established career in Punk band before returning to Michigan to complete her studies in Medicine. It is her life’s vision to create an “audio approach” for sleep study data and using MAX/MSP, our team has already concatenated EEG, EKG, EMG and RespNatal data using EDF formatted files introduced into MAX/MSP and Ableton live. The results sounded like “glitch” meets “Brian Eno/ambient music.” The goal is to create interested music with data that is inherently rather dull and does not change much, but highlight anomalies, such as snoring (or worse), in order to diagnose sleep patients more accurately and efficiently. Sonification of data has been utilized for decades in the medical field but although we can hear the representation of patient data rather ubiquitously in the operating room or intensive care unit (think beeps from EKG and ventilators), sonification as a field of medical research is relatively unexplored in most other specialties.
zzz
UMich Faculty Head +
Mayo Clinic Correspondant
Professor Stephen Rush Dr. Kara Dupuy-McCauley | Pulmonologist
Professor of Performing Arts Technology Mayo Clinic Center for Sleep Medicine
Phase 01
Student Team:
Adam Julian Spencer Harris Omid Shahidi
BS, Computer Science, Sound Engineering BA, Language and Cognition BSE, Computer Science
BFA, Performing Arts Technology
Belany Zhang Bethari Kamarga Logan Gare
BS, Computer Science BS, Computer Science Sound Engineering
BS, Biochemistry Jazz Improvisation
Phase 02
Student Team:
Oliver Zay Eloysa Elena Zelada Cisneros Eilís Finnegan JQ Fan
MS, Applied Statistics MA, Media Arts MArch, Architecture BS, Philosophy
BS, Computer Science
Compositions
Music File 01:
Notice how the beat filters out everytime hypopnea is detected.
Notice how the beat filters out everytime hypopnea is detected.
A Peek into the Process
(insert labeled track corresponding to oxygen drops,
flowchart with screenshots / gifs)
Nasal Pressure Data in the DAW