Pneumonia detection using CNN's algorithm

Electronic instrumentation on medical devices

In today’s world, image processing and machine learning have been the most trending technologies which are used in almost all applications from automation to healthcare. Machine learning (ML) is one of the core fundamentals of Artificial Intelligence (AI) which can solve real life problems. Healthcare is one of the major concerns where AI, ML, Deep Learning (DL) can solve the problems. One of the global healthcare problems is cancer. Cancer refers to a state where abnormal cells develop uncontrollably and affectively damage the healthy body tissue. Cancer often has the ability to spread throughout the body. And among all the cancers, lung cancer was the most common and deadly one. Most lung cancers are detected at a later stage. But if identified at an early stage can save lives. Pneumonia can be a sign of lung cancer and it can be detected using Chest X-ray in most of the cases. For certain applications, the machine learning models can be better or even surpass human professionals. Machine learning and Deep learning technologies play a vital role in predicting the Pneumonia before it gets worse. Hence the idea of this project is to implement a model for Detecting Pneumonia, using different DCNN Algorithms with different combinations of Activation functions and Optimizers, which plays a key role in obtaining the best DCNN model. This work proposes a sequential CNN model, which outperforms the existing standard models like VGG16 and VGG19. The proposed model achieves an accuracy of 95%, precision 93.5, recall 95, f1-score 94.5.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.