The feasibility study of photoplethysmography features for arteriovenous fistula stenosis detection in hemodialysis patients with statistical approach

摘要

Arterial stenosis is highly prevalent in a long-term hemodialysis treatment patient. In this paper, we propose a Rising slope (RS) and Falling slope (FS) which extracted from the right and left-hand photoplethysmography (PPG) signal before and after hemodialysis process. These features ware used to examine 11 patients which could be considered as a method to detect stenosis vessel that occurred after hemodialysis process. The experimental result with a mean + SD age of 77.5 + 12.3 utilizing t-test hypothesis indicated that the variance of RS from left-hand, before and after hemodialysis showed 0.0211 which is appropriate with p <; 0.05 to prove these features statistically significant and can be used as one of proper feature for classification stage.