Pneumonia and Bone Fracture Detection Using CNN

Abstract

Pneumonia is a respiratory infection brought on by bacteria or viruses that affects a lot of people. Pneumonia can result in pleural effusion, a disease where fluids accumulate inside the lungs and make breathing difficult. For effective treatment to be available and to increase survival rates, early identification of pneumonia is crucial. Chest X-ray imaging is the method most frequently used to diagnose pneumonia. Bone fractures are caused by sports, fall, motor vehicle accidents, etc. A radiologist's expertise is also required for the difficult task of promptly and accurately recognizing bone fractures in a huge number of X-ray pictures with a rising number of patients. Accurate bone fracture identification might also be relevant from a medical and legal standpoint. Therefore, a precise diagnosis of the site of bone fractures is crucial in the therapeutic environment. A model that we created from scratch has five layers and is followed by a fully connected neural network. The model was trained using CNN model. The test dataset's accuracy was 98% and 82% for Pneumonia and bone fracture detection respectively.

Ähnliche Arbeiten