Mobile Application

Artificial Intelligence Helps to Reduce Coronavirus Outbreak

Mortality rate in current epidemic (Coronavirus) is close to 3.8% (6500 dead out of 170,000 effected, data till 16March2020). A Race against the clock is underway for the authorities of the countries concerned to manage the outbreaks and try to avoid the spread of the disease. Faced with this challenge, artificial intelligence (AI) is a key tool. AI is a means of anticipating the evolution of the contagion.

Artificial intelligence may well be the solution to the global Covid-19 epidemic, aka the Coronavirus.

As Coronavirus proves increasingly deadly, AI researchers apply machine learning techniques to social media, and other types of data on the web to obtain subtle clues that the disease possibly spreads elsewhere. The new virus appeared in Wuhan, China in December, triggering a global health emergency. Cases of infection and deaths continue to increase.

Can artificial intelligence help cure Coronavirus? The answer to this question is negative at the time, as it is impossible for an AI to create an antibiotic, the Coronavirus not being a bacterium. The situation has changed and not for the better. But science and technology also bring us another evolution. Artificial intelligence is capable of detecting people with Coronavirus in record time.

Different Measures are Taken Through AI to Fight Coronavirus

Diagnosis in 20 Seconds

This feat comes from Alibaba, a Chinese giant in business commerce, and a donation of around 13 million euros allocated to the search for a vaccine against the Coronavirus. While the right medication or drug has not yet been found, the firm has worked on an artificial intelligence system that would diagnose a patient.

See also  A Complete Guide to Develop An Audiobook App Like Audible

The first results are rather conclusive. After a patient has undergone a CT scan, more commonly known as a lung scanner, artificial intelligence will analyze the images obtained and compare them with those of 5,000 other Coronavirus patients. In 20 seconds, AI will be able to determine with 96% reliability whether people are contaminated with the Covid-19 or not.

Social Media and AI

John Brownstein is responsible for innovation at Harvard medical school and expert in the use of data on social networks in the context of emerging health trends. He is part of an international team that uses machine learning to scan publications on social networks, news, data from official public health bodies and information provided by doctors to detect signs that the virus installs in countries other than China.

The AI-integrated program looks for posts on social media that mention specific symptoms, such as breathing problems or fever, in a specific geographic era in which doctors have reported potential cases. Automatic natural language processing (NLP) is used to separate texts published on social networks, for example distinguishing someone who talks about the news and someone who complains about their state of health.

Facial Recognition

Just because artificial intelligence can’t create an antidote to Covid-19 doesn’t mean it won’t help us in the fight against the epidemic. It is Russia that offers a solution, thanks to its gigantic Moscow video surveillance arsenal equipped with facial recognition technology. The Moscow authorities used this device to monitor the 2,500 people deemed to be at risk and to ensure that they respect the quarantine order. This device allows artificial intelligence, with the help of deep learning, to recognize the symptoms of people with Covid-19, especially thanks to their thermal footprint.

See also  5 Data and Security Risks to Watch out for in 2023

Even if people still have to undergo a medical examination, these AI-enabled techniques help diagnose COVID-19. It also allows rapid decision-making to help victims of the Coronavirus and medical professionals to limit the epidemic at best and ensure better care for patients.

    Need an IT Experts? Get a Free Consultation!

    lets start your project

    Related Articles