Lung Respiratory Audio Prediction using Transfer Learning Models
Arohi Patel, Sheshang Degadwala, Dhairya Vyas
Abstract
lung disease is now the leading cause of death worldwide. Lung disease is typically discovered in its final stages, after it has progressed to a serious state. However, the early detection of lung disease may assist in treatment. Technological advancements are critical to the delivery of healthcare services in today’s society. Electronic stethoscopes are used to record the audio clips from patients’ lungs. The audio clips provide useful information for lung diagnosis. The medical community is now focusing on the significance of detecting the lung syndrome using audio acoustics, which is a new research topic. This research study provide a range of transfer learning algorithms for lung audio classification, by utilizing the well-known ALEXNET, VGGNET, and RESNET models. With its high accuracy in classifying the lung audio, the aforementioned transfer learning models might be used to diagnose the lung disorders. Transfer learning strategies, its advantages and disadvantages will be discussed in this research study. Furthermore, this research study also suggests future directions for lung audio identification research as a means of distinguishing between the four distinct lung clips.