Comparative Study on Lung Disease Detection and Classification Techniques using Deep Explainable

One of the main causes of death and disability worldwide is lung illness, commonly known as a respiratory disease. Pneumonia is one of the most common respiratory diseases, and according to the Johns Hopkins Bloomberg School of Public Health report titled “Lung Cancer is one of the leading cause o death among women and men, with a rate of 5 million cases per year. Pneumonia and Diarrhea Progress Report 2020,” 1.23 million children under the age of five died from pneumonia in 2015. These statistics are from the Forum of International Respiratory Societies.334 million people suffered from asthma disease, different pandemics like Covid19 affected lives of many people. Traditional methods for lung detection included blood tests, sputum tests, chest X-rays, and CT scans. More recently, deep learning models have been used to evaluate automatic analysis and classify regions of chest X-rays into diseased and non-diseased states (for example, TB and Non-TB or malignant or benign tumors)[1]. Deep learning’s “black box” model is explained in detail by explainable AI, which also clarifies the patient’s or examination’s symptoms. In this paper we will exaggerate different deep learning and explainable techniques along with parameter comparison for diagnosis of pulmonary diseases like TB, lung cancer, pneumonia and interstitial lung diseases. This survey aims at getting acquainted with different techniques of detection of lung disease’s using various types of CNN model’s and various types of machine learning classifiers with explainable AI like Grad CAM, Score CAM etc for visualization of disease infected regions Also the study and analysis of different models will help researchers in their studies. Future scope and challenges are also given for further expansion of techniques.