• Version
• Download 13
• File Size 266.30 KB
• File Count 1
• Create Date 14 April 2023
• Last Updated 14 April 2023

Study of Chest X-rays to Analyse Thorax Diseases using Deep Learning Framework

Sudarshan Adhav¹, Rohit Borse², Riya Gujarathi³, Madhuri Kubade⁴, Prof. Priyanka Khalate⁵

¹²³⁴⁵Department of Computer Engineering, Smt. Kashibai Navale College of Engineering Vadgaon(Bk.), Pune

Affiliated to Savitribai Phule Pune University

Abstract— When a patient's doctor suspects they have a lung ailment, one of the first tests they frequently have is a chest X-ray. Sometimes vital information regarding the sickness is missed by radiologists or restorative inspectors. It takes a lot of time to find the sickness' manifestations when these x-beams are reexamined. In this study, we demonstrate how to diagnose thoracic problems using a deep convolutional neural network (CNN). We begin by matching the interest points on the photos to align them. The dataset is then expanded using the theory of Gaussian scale space. The larger dataset is then used to train a deep CNN model in the following step. The larger dataset is then used to train a deep CNN model in the following step. Then, more recent test data that we have are diagnosed using the model. Our research demonstrates how effective our strategy is at producing quality outcomes.

Keywords–image classification, machine learning, imageprocessing, neural network.