Arduino Nano 33 BLE Sense Performance for Cough Detection by Using NN Classifier

Coughing is the body's natural response to expelling substances and particles from the airways to prevent entry into the lower airways. As the advancement of Artificial intelligence technology has now applied in various fields. One of them is in the field of artificial health intelligence which is currently widely used to recognize various symptoms disease. In this research, a cough detection system will be designed and made by using cough sound data obtained by using Arduino 33 BLE Sense and Edge Impulse for development. Which means we can do classification without building, deploying and scalling embedded Machine Learning Application easier and faster. Arduino Nano 33 BLE Sense consists of multiple built-in sensors such as 9-axis inertial, humidity, temperature, barometric, microphone, gesture, proximity, light color and light intensity sensors. The cough sound detected by the microphone of Arduino Nano 33 BLE Sense with 57 total of cough sound then processed for classification using machine learning edge impulse tools. The cough sample sounds are given 4 types of labels, namely cough (batuk), cough with noise (batuk bising), silence (hening), Noisy (bising) and the result of classification using neural network in Edge impluse are 73,6% for accuracy and 0.86 for loss.