Atrial Fibrillation (AF)
Conditions
Keywords
Smartphone, Persistent Atrial Fibrillation, Mobile Health, mHealth, Telemedicine, Artificial Intelligence, Heart Sounds, Phonocardiography, Photoplethysmography, Electrocardiography, 12-Lead ECG, 12-Lead EKG, 12-Lead Electrocardiography, ECG, EKG, Electrocardiogram, Electrocardiograph
Brief summary
The purpose of this 4-in-1 observational study is to test the performance of artificial intelligences (AIs) in distinguishing irregularly irregular heart rhythm called atrial fibrillation (AF) from normal heart rhythm using physiological signals collected by smartphones' built-in hardware and/or external accessories. Participants will: * Have their weight, height, resting heart rate and blood pressures measured * Have 12-lead electrocardiogram (ECG) of their heart electrical activities recorded * Have their heart sounds and 1-lead ECG recorded from their chest, and optical-based blood flow data (photoplethysmography or PPG) and 1-lead ECG recorded from their fingers using smartphones' built-in microphone, camera, and/or external accessories * Optionally have their optical-based blood flow data recorded from their face using smartphones' built-in camera (remote PPG or rPPG). The researchers will also create a database containing the physiological signals collected in this study along with the participants' medically relevant information to help train and test future AIs for medical applications.
Detailed description
4 observational studies have been combined into 1 observational study to share the same pool of participants. These 4 studies are designated as AUSC-AF, ECG-AF, AUSC+ECG-AF, and rPPG-AF corresponding to the signal modality/modalities used for AF detection (see outcome measures) and designated as AUSC/ECG/rPPG-AF when combined as 1 study.
Interventions
DIAGNOSTIC_TESTComputer algorithms
Computer algorithms that are designed to perform heart sound, electrocardiography (ECG), and/or facial photoplethysmography (rPPG) analysis on data collected from smartphone's internal hardware and/or external accessories.
Sponsors
Laboratory of Data Discovery for Health
The University of Hong Kong
Study design
Observational model
CASE_CONTROL
Time perspective
CROSS_SECTIONAL
Eligibility
Sex/Gender
ALL
Age
22 Years to No maximum
Inclusion criteria
* Age: ≥22 years (adult) * Patients who have one of the following: * Permanent atrial fibrillation, or * Long-standing persistent atrial fibrillation (12 months or longer), or * Confirmed 12-lead ECG diagnosis for persistent atrial fibrillation (\> 7 days) or sinus rhythm within 12 months at the time of their normal attendance at the hospital
Exclusion criteria
Any of the following: * Implanted active medical devices in the torso, such as pacemakers and defibrillators * Patients without atrial fibrillation who have another arrhythmia * Completely missing one or more limbs, or missing any hand * Disability in using their hands or arms * Lack of both index fingers, or all fingers in any hand * Both index fingers with any of the following characteristics: * Tattooed/inked * Reduced blood flow in the fingertip (e.g. perniosis or callus formation)
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Differentiation of Atrial Fibrillation from Sinus Rhythm in Heart Sound Recordings | Day 0 | AUSC-AF: Identification of atrial fibrillation (AF) from sinus rhythm in recorded heart sounds (phonocardiogram \[PCG\]) as verified by the gold standard 12-lead electrocardiography (ECG) interpretation, measured in the form of sensitivity and specificity. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Differentiation of Atrial Fibrillation from Sinus Rhythm in Heart Sound Recordings | Day 0 | AUSC-AF: Identification of atrial fibrillation from sinus rhythm in recorded heart sounds (phonocardiogram \[PCG\]) as verified by the gold standard 12-lead ECG interpretation, measured in the form of positive and negative predictive values, and accuracy. |
| Differentiation of Atrial Fibrillation from Sinus Rhythm in 1-Lead ECG Signals | Day 0 | ECG-AF: Identification of atrial fibrillation from sinus rhythm in recorded 1-lead ECG signals as verified by the gold standard 12-lead ECG interpretation, measured in the form of sensitivity, specificity, positive and negative predictive values, and accuracy. |
| Differentiation of Atrial Fibrillation from Sinus Rhythm in PCG and 1-Lead ECG Signals | Day 0 | AUSC+ECG-AF: Identification of atrial fibrillation from sinus rhythm in PCG and 1-lead ECG signals as verified by the gold standard 12-lead ECG interpretation, measured in the form of sensitivity, specificity, positive and negative predictive values, and accuracy. |
| Differentiation of Atrial Fibrillation from Sinus Rhythm in Facial Photoplethysmography Signals | Day 0 | rPPG-AF: Identification of atrial fibrillation from sinus rhythm in facial photoplethysmography signals (also known as remote photoplethysmography \[rPPG\]) as verified by the gold standard 12-lead ECG interpretation, measured in the form of sensitivity, specificity, positive and negative predictive values, and accuracy. |
Countries
China
Outcome results
None listed