New approach of statistical analysis for lung disease diagnosis using microscopy images

Lungs play an important role in human respiratory system. There are various diseases that effect the functioning of the lungs. Diagnosis and treatment of the disease is most important for proper functioning of human body. The most difficult part is to detect and diagnose the diseases at primary stage. In this paper, a computerized method for lung disease diagnosis is implemented. The lung diseases like Fibrosis, Emphysema and lung Cancer is detected using MATLAB software as these diseases are having almost similar symptoms and hence difficult to diagnose. The algorithm implemented here is based on Texture analysis, Principle Components Analysis and various image quality measurement statistical parameters such as Entropy, Standard Deviation and Texture index. Fuzzy logic is used to calculate the performance parameters of lung images. The algorithm is implemented and tested on microscopy lung images