| Summary: | Nap is an effective way to reduce daily-level fatigue after several hours of work. However, no alarm clock, which intelligently manages the nap duration with good autonomic nervous recovery (ANR) from fatigue, has been reported in literature. In this work, an intelligent biological alarm clock algorithm was designed on the basis of electrocardiogram (ECG) and electroencephalogram (EEG) data acquisition and analysis. ECG data samples were collected from 31 subjects in 278 times of nap experiments and categorized into good, moderate, and poor ANR datasets according to the degree of sympathetic withdrawal and parasympathetic activation during the nap. In practice, the alarm clock automatically classified the new-coming ECG data as good, moderate, or poor ANR with a classifier trained by the abovementioned ANR datasets. A prototype system of the intelligent alarm clock algorithm was implemented and validated in real-scene naps. The prototype system detected falling asleep during the closed-eye naps with a true positive rate of 93.55% and a true negative rate of 100%.
|
|---|