Motion and Noise Artifact-Resilient Atrial Fibrillation Detection Using a Smartphone

Smartphone signals corrupted by motion and noise artifacts (MNAs) are often misclassified into atrial fibrillation (AF) by our previous smartphone AF detection application. We developed an MNA-tolerant AF detection algorithm for smartphones, which first detects MNAs in the smartphone signals, removes them, and finally detects AF from the MNA-free smartphone signals. To detect MNAs, we used time and frequency-domain parameters: high-pass filtered signal amplitude, successive pulse amplitude ratio, and successive maximum dominant frequency. AFs are detected using our previous AF detection algorithm based on root mean square of successive RR difference (RMSSD) and Shannon Entropy (ShE) values. The clinical results show that the accuracy, sensitivity and specificity of the proposed AF algorithm are 0.9632, 0.9341, and 0.9899, respectively.