We are actively engaged in research to create new technology for collecting physiological information, and machine learning-based algorithms for automated detection of disease conditions. This page describes some of our ongoing research projects and relevant peer-reviewed scientific publications.

Can we use machine learning + mobile technologies to prevent strokes?

Screening/Detecting Atrial Fibrillation from Optical Signals (PPG)

Atrial fibrillation (AF), the most common heart rhythm disorder, produces an irregular heartbeat and is a leading cause of all strokes. Many of these AF-related strokes are preventable through medication if a timely diagnosis is made. However, because AF often produces vague or no symptoms, AF often goes undetected.

We are developing a new, patent-pending technology, Cardiio RhythmTM, that can detect the presence of AF from optical signals collected using a smartphone camera, or wearable devices/smartwatches (for example, an Apple Watch or Fitbit) that use photoplethysmography (PPG). Cardiio RhythmTM uses a machine learning approach to identify the signature pattern of AF and distinguish it from other heart rhythms. We are collaborating with clinicians from Massachusetts General Hospital, Hong Kong University, and Chinese University of Hong Kong to validate the performance of this technology. Some of our published results are listed below.

Cardiio RhythmTM is not part of the "Cardiio-Heart Rate Monitor + 7 Minute Workout" app that is currently available on the App Store for non-medical use only. Cardiio RhythmTM is not FDA-cleared and is not for sale.

Clinical Validation of Cardiio RhythmTM Technology for Detecting Atrial Fibrillation

Validation of a Novel Contact-free Atrial Fibrillation Screening Method Using an Iphone Camera to Detect Facial Pulsatile Photoplethysmographic Signals
Circulation, 134.Suppl 1 (2016)
Authors: Yan BP, Chan S, Lai WH, Lau BH, Lam EK, Yip AK, Tai NL, Ng OC, Chan HL, Poh YC and Poh MZ
Read More

Diagnostic Performance Of A Smartphone-Based Photoplethysmographic Application For Atrial Fibrillation Screening In A Primary Care Setting.
Journal of the American Heart Association, 5(7), p.e003428 (2016)
Authors: Chan PH, Wong CK, Poh YC, Pun L, Leung WWC, Wong YF, Wong MMY, Poh MZ, Chu DWS and Siu CW
Read More

Head-To-Head Comparison Of A Camera-Based Smartphone Application Cardiio Rhythm With Alivecor Heart Monitor For Atrial Fibrillation Screening In Primary Healthcare Setting
Journal of the American College of Cardiology, 67.13_S (2016)
Authors: Siu D, Wong CK, Chan PH, Poh YC, & Poh MZ
Read More

Diagnostic Accuracy of a Novel Mobile Application (Cardiio Rhythm) for Detecting Atrial Fibrillation
Journal of the American College of Cardiology, 65.10_S (2015)
Authors: Vaid J, Poh MZ, Saleh A, Kalantarian S, Poh YC, Rafael A, & Ruskin J
Read More

Validation of Cardiio's Heart Rate Measurement Technology

We validated the accuracy of the "Cardiio - Heart Rate Monitor + 7 Minute Workout" mobile app on people across a range of skin tones, and both at rest and after performing exercises at different intensity levels.

Validation of the "Cardiio - Heart Rate Monitor + 7 Minute Workout" Mobile App for Measuring Heart Rate

Resting and Postexercise Heart Rate Detection From Fingertip and Facial Photoplethysmography Using a Smartphone Camera: A Validation Study
JMIR Mhealth Uhealth, 5(3): e33 (2017)
Authors: Yan BP, Chan CK, Li CK, To OT, Lai WH, Tse G, Poh YC, Poh MZ
Read More

Validation of a Standalone Smartphone Application for Measuring Heart Rate Using Imaging Photoplethysmography
Telemedicine and e-Health, 23(8), p678-683 (2017)
Authors: Poh MZ and Poh YC
Read More