| Title |
Photoplethysmography-based system for atrial fibrillation detection during hemodialysis / |
| Translation of Title |
Fotopletizmografija pagrįsta sistema prieširdžių virpėjimui atpažinti hemodializės procedūrų metu. |
| Authors |
Stankevičius, Dainius |
| Full Text |
|
| Pages |
63 |
| Keywords [eng] |
photoplethysmography ; atrial fibrillation ; hemodialysis ; photoplethysmogram quality assessment ; unobtrusive system |
| Abstract [eng] |
Atrial fibrillation is a relevant problem among patients undergoing hemodialysis. In this work, we propose an unobtrusive system for atrial fibrillation detection. The system consists of a wearable device and integrated algorithm for real-time atrial fibrillation detection using solely the photoplethysmogram. The algorithm is based on the heart rate irregularity and consists of low-pass filter for high-frequency noise removal, adaptive filter for baseline removal, slope sum function and adaptive threshold for peak detection, low-complexity atrial fibrillation detector for decision making, and mean-crossing rate based measure for signal quality assessment. When testing the performance of the algorithm with the collected set of noise-free signals, the sensitivity of 0.99 and the perfect specificity of 1 were obtained. The performance of the proposed low-quality-signal detector showed a sensitivity of 0.99 and a specificity of 0.90 for a distinction between noise-free photoplethysmogram and photoplethysmogram contaminated with added noise during both - atrial fibrillation and normal rhythms. |
| Dissertation Institution |
Kauno technologijos universitetas. |
| Type |
Master thesis |
| Language |
English |
| Publication date |
2016 |
