Advanced search
Start date
Betweenand

Chest X-ray image classification using deep neural networks

Grant number: 19/20875-8
Support type:Scholarships in Brazil - Master
Effective date (Start): January 01, 2020
Effective date (End): February 28, 2021
Field of knowledge:Physical Sciences and Mathematics - Computer Science
Principal researcher:Zanoni Dias
Grantee:Vinicius Teixeira de Melo
Home Institution: Instituto de Computação (IC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil

Abstract

Chest X-ray is one of the most common radiology exams, being used, for example, to diagnose and monitor the treatment of various pulmonary conditions, such as pneumonia, emphysema and cancer. Automated identification and interpretation of this type of exam, at the same level or better than radiologists, could enable advances in a variety of medical environments, from improved outcome speed to global health initiatives. In this project, we propose a method for chest X-ray images classification based on Deep Neural Network architectures. To validate the developed models, the experiments will be performed using the ChestX-ray14 and CheXpert datasets. The methodology used in this project is based on extraction of activation maps of the neural network to crop the images considering only the areas of interest, discarding irrelevant information for their classification. (AU)

News published in Agência FAPESP Newsletter about the scholarship: