Detection of Various Lung Diseases using Machine Learning

Abstract

There are many applications of machine learning in various fields among which  diagnosis of diseases in early stages proves to be one of the important application in the medical field. Image processing techniques plays important role in various disease detection. Among the various diseases, lung diseases are most harmful diseases with more than 225,000 peoples getting affected with it and causing death of almost 150,000 peoples. Detection of lung diseases at a very early stage can be useful specially to improve the survival rate of a patient. The aim of this research is to detect bronchiectasis, lung cancer, hanta virus and pneumonia using CT scan image. In this research a quick and well balanced system of lung disease detection is proposed. In contradiction to previous studies, fully connected network requires more number of neurons in first hidden layer of artificial neural network while CNN algorithm requires less number of neurons.