How a 19th-Century Breakthrough Could Revolutionize COVID-19 Diagnosis

To treat someone you suspect of having coronaviruses, you need to confirm that they are actually carrying the disease.

In the UK, it is easy to overlook the fact that we have had a reliable method for detecting patients with infections since the beginning of the Pandemic. This allowed for people to be treated and cared for quickly.

The main method for determining if someone has COVID is reverse transcription polymerase chain reaction testing. The coronaviruses genetic material can be found in a sample taken from a person.

It is a major challenge to do this testing at scale because of the resources needed.

High-income countries have been able to scale up their COVID testing at great cost, but in some low- and middle-income countries, health staff have not been able to do so. This is a problem in remote areas.

On top of that, the testing is very slow. It takes around two hours to get the sample to a lab for testing.

In many cases, it is necessary to confirm whether someone has the virus quickly. Treatment needs to start immediately when someone has severe COVID. It is possible to quickly diagnose the disease.

Our team looked into the possibility of providing a quick and reliable alternative to PCR testing by using machines in the radiography department.

Diagnosticians can use computed tomography or X-ray techniques to look for signs of a COVID infection.

The World Health Organization recommends using radiography for detecting COVID in patients who aren't able to get a test for the disease because of the lack of available testing.

There is a resource shortage here as well. Colovid's visual pointers can be hard to spot, so using X-rays andCT scans for diagnosis requires radiologists to carefully decipher the chest images. An artificial intelligence program was created to speed up diagnosis and allow radiologists to work.

The program is based on a type of neural network that is used to analyze images. The key features of an image can be picked out by the program.

We began by training and testing a number of different algorithms using a database of around 3000 chest X-rays. There were scans from patients with COVID, healthy individuals and people with viral pneumonia.

We made the algorithm better at spotting the differences between the X-rays. One clearly performed better than the others.

We gave this top performer a completely new set of X-rays that it hadn't seen before, and asked it to determine if each came from a COVID patient or not. 98% of the time, the program got the answer right.

How could this be used?

We developed an app that could be used outside of our lab to make a difference, so that it could be used in places where it could make a difference. The app can be installed on normal PCs and laptops without the need for a lot of computer memory or power.

It has been designed in such a way that there is no need for additional equipment. Patient X-rays can be uploaded to the app via the web, and then the app analyses the image and gives a result indicating if it is positive or negative.

This app won't replace the PCR. In A&E departments, it could be very effective. It would allow for a chest X-ray to be taken and analyzed quickly, and if the patient is positive, for treatment to start immediately rather than waiting for lab results.

As well as being beneficial for patients, this could also speed up their passage onto suitable wards elsewhere in the hospital, and so relieve the strain on hard-pressed A&E departments.

In low-income countries and remote areas, the app could be very effective at detecting COVID cases. We are going to test it out in Pakistan as part of the EU-funded SAFE RH project to see what impact it can have in the real world.

Naeem Ramzan is the Professor of Computing Engineering at the University of the West of Scotland.

This article is free to use under a Creative Commons license. The original article is worth a read.