Identifying COVID-19 patients from CT images

COVID-19 is an infectious disease caused by a virus leading to severe acute respiratory syndrome corona virus (SARS-CoV) which has taken the form of a pandemic. In the face of the pandemic, there is a need for fast and accurate diagnosis of COVID-19 positive cases. This project aims at identifying and screening suspected patients as COVID-19 positive or negative using CT images (dataset: https://github.com/UCSD-AI4H/COVID-CT or any other appropriate dataset), for further treatment without having to wait for the results of pathogenic tests. The main tasks in the project would be as follows:

1. Identifying areas of consolidations in the image,

2. Binary Classification of the patients to identify COVID-19 positive cases

3. Determine the severity of the infection.

This project involves the use of image processing (enhancement, segmentation, reconstruction and image analysis) and machine learning (classification) techniques.

Supervisor name: 
Dr. Gule Saman
Supervisor and Deputy email addresses: 
s.gule@hw.ac.uk