Using AI To Combat SARS COVID-19

The current global epidemic caused by the SARS Covid-19 virus has got countries around the world scrambling to find new ways to deal with the fallout from it. From the logistics of testing and diagnosing patients the vast number of patients on a daily basis, to identifying the various reported strains of the virus in order to expedite the process of designing a vaccine for them, there are a lot of areas where the need for faster, automated systems is apparent.

Countries around the world had very different responses and approaches to combatting the disease, with those who did not take sufficient measures to curb the rate of infection in the early days now having patients in the hundreds of thousands, as well as a daily infection rate in the tens of thousands. The response teams around the world are trying hard to catch up with virus, but it is getting harder and harder to catch up with every passing day. The situation for many countries is so critical that experts predict that this disease may still be a danger for about another two years!

Data scientists have been a vital resource in helping authorities work through the large amount of data generated during this epidemic, such as the helping diagnose new cases using insights gained from previous patient data, or helping virologists test the virus’s response to various antivirals and other medicines in order to find a viable treatment plan. But devising processes that can help speed up these processes, or ideally automate them, are needed urgently.

Enter Artificial Intelligence:

AI solutions are the way to go forward combating this epidemic. We have the data, but no time to evaluate it, and an acute shortage of manpower to boot. Many research teams around the world have been working on AI services to help the people working on the frontlines. These innovative solutions are now helping bridge the gap between the disease and us. Some of the most cutting edge solutions are:

• Alibaba’s AI based Covid-19 diagnostic system Using Chest CT
• Insightiv’s Detection of suspected Covid-19 patients in Rwanda using Chest CT and X-rays
• AI based surveillance systems designed to detect Covid-19 patients early
• AI chatbots designed to combat misinformation about Covid-19

There are many other AI based solutions which are still under work, but for now we will discuss the solutions which are already operational.

Alibaba’s AI based Covid-19 diagnostic system Using Chest CT:

In response to the problems medical teams faced checking medical reports of Covid-19 in China and around the world, Alibaba’s R&D team launched an AI based system capable of diagnosing the presence of Covid-19 virus using patients’ CT scans of the chest. They claim the system can detect the presence of the disease in less than twenty seconds, with an accuracy of 96%. Compared to the AI system, manually working through the CT images can take a medical professional approximately 15 minutes, with a lower accuracy rate as the volume of data increases. More than a hundred hospitals have benefited from the integration of their AI model into their diagnostic process for faster diagnostic speeds.

Insightiv’s Detection of suspected Covid-19 patients in Rwanda using Chest CT and X-rays:

The team of researchers and engineers from Insightiv have designed an AI based diagnostic system that can detect the presence of corona virus in the patient using CT scans and X-ray of the chest. The system had already been under development by the team in order to facilitate teleradiology, which is the sharing of radiological patient reports in order to share studies with doctors who are geographically far, in Rwanda and other African countries. When the covid-19 pandemic broke out, they fast-tracked their development into building a system that would flag a suspected covid-19 patient based on their radiology reports, and bring it immediately to the radiologist’s attention so that they can verify the detection and report it to the relevant authorities.

This solution is an efficient way to speed up the process for the medical personnel, while not being technologically expensive, so that deployment of the system is possible in areas which are less financially capable.

AI based surveillance systems designed to detect Covid-19 patients early:

As countries like South Korea and China proved, tracking potential patients using people’s phone records, bank activity and the like, a country can curb the spread of such infectious diseases. While these actions may give rise to issues of personal privacy and other civil liberties, using less violating methods such as tracking and identifying people not wearing protective masks or not following social distancing rules can be quite beneficial into conditioning people to follow the rules set in place for their safety.

AI chatbots designed to combat misinformation about Covid-19:

This pandemic has resulted in a lot of wrong, downright harmful information floating around, which only serves to cause increased levels of panic and hysteria amongst the masses. Not only that, people following the wrong information are not only hurting themselves, but others around them as well. Patients trying to cure themselves using quack remedies, or believing that the Covid-19 virus is due to the 5G network testing and going out and about without masks and proper precautions are only contributing to the continued spread of the disease.

To counter that, many health services have developed AI based chatbots which are available to answer the queries people may have about the entire situation. These chatbots learns what the people generally want to know using the various verified FAQ pages, and updates its responses based on the queries people ask of it. These chatbots are also regularly verified by various medical professionals to check the quality of the information being given.

To summarize, many people are designing an AI strategy to speed up or solve the problems caused by the Covid-19. While some are successful, others may need a bit more work. But what the plans which were successful have shown us now is that using AI, we can finish tasks which would be impossible, or extremely hard at the very least.