Atrial fibrillation detection using photo-plethysmography and acceleration data at the wrist

Atrial fibrillation (AF) is a pathological cardiac
condition leading to increased risk for embolic stroke.
Screening for AF is challenging due to the paroxysmal and
asymptomatic nature of the condition. We aimed to
investigate whether an unobtrusive wrist-wearable device
equipped with a photo-plethysmographic (PPG) and
acceleration sensor could detect AF. Sixteen patients with
suspected AF were monitored for 24 hours in an outpatient
setting using a Holter ECG. Simultaneously, PPG and
acceleration data were collected at the wrist. PPG data
was processed to determine the timing of heartbeats and
inter-beat-interval (IBI). Wrist acceleration and PPG
morphology were used to discard IBIs in presence of
motion artefacts. An ECG validated first-order Markov
model was used to assess the probability of irregular
rhythm due to AF using PPG-derived IBIs. The AF
detection algorithm was compared with clinical
adjudications of AF episodes after review of the ECG
records. AF detection was achieved with 97 ± 2%
sensitivity and 99 ± 3% specificity. Due to motion
artefacts, the algorithm did not provide AF classification
for an average of 36 ± 9% of the 24 hours monitoring. We
concluded that a wrist-wearable device equipped with a
PPG and acceleration sensor can accurately detect rhythm
irregularities caused by AF in daily life.