Insomnia is a sleeping sickness in which the time and quality of human sleep is not sufficient due to difficulty getting to sleep or difficulty maintaining sleep. Early diagnosis and treatment of insomnia is necessary to prevent chronicity and death resulting from untreated insomnia. The body's required ECG signal can be detected by installing a sensor on the human body. The ECG signal has several points, namely P, Q, R, S and T points. There are PR intervals, PR segments, QRS complex, ST segments, and QT intervals as areas in the ECG signal. In this study, we will use the R-R interval of human ECG signals taken when the sleep condition is for 2 hours which already represents one person's sleep cycle. The features to be used are the Mean RR and SDNN features. The K-Nearest Neighbours method classifies new data by finding the shortest distance from the training data, this makes this method suitable for this study because the training data used has significant differences between one class and another class. The tools that will be used to detect insomnia are the Arduino Uno microcontroller and the AD8232 module which are used to detect and filter the detected signals. AD8232 is a module used to acquire ECG signals. The use of K-Nearest Neighbours as a classification method has an accuracy of 86% for K = 3 and obtain an average computation time of 79.9 ms.
Copyrights © 2023