Lung Sounds Identification based On Transfer Learning Approaches : A Review

“lung diseases are now considered as one of the fatal diseases across the globe. However, early detection of lung disease may help in providing earlier treatment since most cases of lung diseases are only detected after they have progressed to advanced stage. Today's healthcare system relies on the recent technological advancements. Lung sound analysis plays a crucial role in the diagnosis of lung disease. Further, the successful navigation of medical system requires the ability to acquire new information and utilize it in new contexts. To perform classification, this research work presents several transfer learning strategies, including ALEXNET, VGGNET, and RES NET for analyzing the lung sounds. To complement the techniques, a Transfer learning model that incorporates a Modified RESNET with a Mel spectrogram of lung sound signals are used to perform classification. These transfer learning models perform efficiently in classifying the lung sounds, which can be later used to diagnose respiratory diseases. This research study analyzes several transfer learning methods and discuss their benefits and drawbacks in identifying four distinct types of lung sounds. Finally, the further research directions on the identification of lung sounds are discussed.”