Lung Sound Recognition Using Spectrogram and Adaptive Resonance Theory 2 Neural Network (ART2)

Usually, physicians diagnose lung diseases by listening to the lung sound using stethoscope. This technique is known as auscultation. Some lung diseases produce unique lung sounds, which refer to special recognized pattern. But the main problems concerning are the lung sounds that have low frequency (20 – 2000 Hz), low amplitude, in addition to other factors such as interference from other sounds, ear sensitiveness, and low variety of the pattern of lung sounds that make them almost similar. These came factors lead to the false diagnosing of lung disease if the auscultation procedures are not conducted correctly. We proposed method to classify lungsound using spectrogram and ART2 neural network. By this method we can classify lungsound abnormality based on peak frequency each time. Experimental result shows accuracy of the system up to 98% for 70 testing data that consist of 5 classes of lung sound abnormality.