A graduate course final project

This project proposes a novel convolutional neural network model that was trained from the ground up to classify and detect the presence of Pneumonia in a set of chest X-ray image samples. Unlike other methods that rely solely on transfer learning approaches or traditional handcrafted techniques to achieve remarkable classification performance, a convolutional neural network model was built from the ground up to extract features from a given chest X-ray image and classify it to determine if a person is infected with pneumonia. This model could help mitigate the reliability and interpretability challenges that are frequently encountered when dealing with medical imagery. This project was created based on one of the substantial papers which had been published in the Hindawi journal, which the name is “An Efficient Deep Learning Approach to Pneumonia Classification in Healthcare” (Stephen, Sain, Maduh, & Jeong, 2019). The novel architecture of the neural network was successful enough to have a better performance in comparison to the proposed CNN by the paper. In this project, the steps and the differences between the two CNN was highlighted.

• Histogram of X-ray photos having Pneumonia infection and not:

The invented CNN architecture (First plot) for classifying x-ray photos is compared with the proposed CNN of the paper (second ploy):

The performance of the invented CNN and proposed CNN on test data:

Confusion matrix for invented CNN for investigating errors:

The video of the presenting my project: