Diagnostix - COVID-19 Identification through Radiographic Images
Shriram KV
Bengaluru, Karnataka
We all are aware that One major symptom of COVID-19 is Pneumonia. The challenge lies in differentiating common cold from COVID-19. So, we propose a tool for the identification of COVID-19 through Deep Learning-based Classification of Radiographic Images. It works as follows: 1. The X-ray image of t ...learn more
Project status: Published/In Market
Intel Technologies
Intel NUC
Overview / Usage
One of the main symptoms of covid19 is pneumonia. We propose a Deep Neural Network Transfer learning approach to identify if a patient is affected by pneumonia and further identify if it is COVID pneumonia or non-COVID pneumonia. Our web/mobile application gets input and processes the input with a multi-level deep learning architecture which includes a custom-tuned Convolutional Neural Network at the first level to identify the presence of pneumonia and the second level to identify if the existing pneumonia is COVID-pneumonia or non-COVID pneumonia and displays the identified result in the same web/mobile application. The key objectives of the solution include:
-
Pneumonia detection caused by covid19
-
Reduce load on doctors
-
Reduced time of identification of probable candidates
-
Reduction of manpower.
-
Tests only on most probable suspects
Methodology / Approach
A simple UI where the Doctors can upload the X Ray Images. In few seconds the result shall be presented as Normal, Pneumonia (Non Covid) or Covid 19 +
Technologies Used
Angular,
Deep Learning,
Python