Grant: Master's Award: Canada Graduate Scholarships
University of Toronto
$17,500.00
Development of Deep Learning Models for Motion Artifact Reduction in Wearable Cardiac Monitoring Devices
EF 2023-24
Résumé
Wearable devices are assuming a larger role in remote healthcare, fitness tracking, and athletics. Predominantly based on photoplethysmography (PPG), these devices use light to non-invasively detect changes in blood flow and oxygenation within peripheral circulation - leading to estimates of heart rate, pulse oximetry, and the identification of arrhythmias. It's projected that future wearables combined with AI algorithms will revolutionize the quality of care for patients with heart failure and cardiovascular disease by obtaining remote datasets over timescales and locations previously unattainable. However, the presence of motion greatly reduces the quality and interpretability of data from wearable devices and limits the development of diagnostic AI models. Currently, accelerometers are used to detect excessive motion and dispose of 'contaminated' data. Attempts to recover PPG data using accelerometer data and AI has been challenging, primarily because accelerometers capture global motion, not the relative motion at the sensor-skin interface - the predominant source of motion artifacts. Here, we propose the reconstruction of physiological waveforms in the presence of motion through the development of a multimodal sensor and deep learning model. The sensor will combine force and multiwavelength optical measurements to capture relative motion at the sensor interface. Deep learning models will then be trained to reconstruct the PPG waveform. Initial testing and training will consist of controlled lab experiments and then translate into real-world examples of motion. This work will enable real-time motion artifact cancellation in wearable optical devices and lead to more advanced remote health care devices, athletic performance trackers, and algorithms.
Programme
Master's Award: Canada Graduate Scholarships
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
CIHR Funding Decisions Database
Tel que publié
Id
506372
Name
Lee Matthew K
Region
Ontario
Country
Canada
Orgtype
Institution Paid; Research Institution
Prcname2
Special Cases - Awards Programs
Projecttitle
Development of Deep Learning Models for Motion Artifact Reduction in Wearable Cardiac Monitoring Devices
Programtype2
Studentship Programs
Orgnameinp2
University of Toronto
Cihramount2
$17,500
Cihrequipment2
$0
Programname2
Master's Award: Canada Graduate Scholarships
Keyworddelim
Artificial Intelligence; Cardiac Vital Estimation; Cardiology; Deep Learning; Motion Artifacts; Photoplethysmography; Remote Cardiac Monitoring; Signal Reconstruction; Supervised Learning; Wearable Devices
Abstract
Wearable devices are assuming a larger role in remote healthcare, fitness tracking, and athletics. Predominantly based on photoplethysmography (PPG), these devices use light to non-invasively detect changes in blood flow and oxygenation within peripheral circulation - leading to estimates of heart rate, pulse oximetry, and the identification of arrhythmias. It's projected that future wearables combined with AI algorithms will revolutionize the quality of care for patients with heart failure and cardiovascular disease by obtaining remote datasets over timescales and locations previously unattainable. However, the presence of motion greatly reduces the quality and interpretability of data from wearable devices and limits the development of diagnostic AI models. Currently, accelerometers are used to detect excessive motion and dispose of 'contaminated' data. Attempts to recover PPG data using accelerometer data and AI has been challenging, primarily because accelerometers capture global motion, not the relative motion at the sensor-skin interface - the predominant source of motion artifacts. Here, we propose the reconstruction of physiological waveforms in the presence of motion through the development of a multimodal sensor and deep learning model. The sensor will combine force and multiwavelength optical measurements to capture relative motion at the sensor interface. Deep learning models will then be trained to reconstruct the PPG waveform. Initial testing and training will consist of controlled lab experiments and then translate into real-world examples of motion. This work will enable real-time motion artifact cancellation in wearable optical devices and lead to more advanced remote health care devices, athletic performance trackers, and algorithms.
Competitiondate
202312
Approvedterm2
1 yr 0 mth
Pinamesdelim
Lee, Matthew K
Approvedterm
1 an 0 mois
Cihrcontribution2
~ $17,500~ $0~