Utilizing Convolution Neural Networks for the Aucoustic Detection of Inhaler Actuations
Asthma is a chronic respiratory disease and a significant burden for patients, their families and the healthcare system as a whole. Unfortunately, the management of the disease is far from optimal mainly due to the reduced adherence of patients to their medication plan. In order to solve this problem, a number of novel inhalers have been proposed over the past that monitor and support the proper use of inhaled medication. Aiming in this direction, the current study investigates the use of acoustic signals for the detection of inhaler actuations during activities of daily living and outside the controlled environment of the laboratory. The proposed algorithm is based on Convolution Neural Networks. The results of the current approach, have led to high levels of accuracy (98%), demonstrating the potential of this method for the development of novel inhalers and medical devices in the area of respiratory medicine.
D. Kikidis, K. Votis, D. Tzovaras. "Utilizing Convolution Neural Networks for the Aucoustic Detection of Inhaler Actuations". In Proceedings of the IEEE International Conference on E-Health and Bioengineering. Iasi, Romania, 19-21 November 2015.