AI/ML approaches to aid COVID-19 combat

As the total number of COVID-19 reported cases reaches 1,358,469 as of 7 April 2020. And countries like Spain & Iran reported 3835 & 2089 number of reported in one day, countries across the globe are trying their best to combat this crisis. Several countries including The US, China, South Korea, Taiwan are implementing AI-based solutions to combat COVID-19.

AI can not only predict the start of an epidemic but can also forecast how it will spread. Quick response, prediction, and prevention help stalling outbreaks of epidemics such as the COVID-19.  Natural Language Processing, Machine Learning, and genetic applied science can successfully study travel patterns, traveler data, health organizations data, climate data from satellites and more to provide new insights to address the outbreak and help contain the devastating effects.

The role of AI becomes more relevant to contain the spread of the disease COVID-19 because of its technological developments and innovative techniques.  AI helps in predicting the flow of the virus by drilling deep into the data collected.

So, what kind of data is needed for AI to predict the flow and pattern of COVID-19 across the countries? Data such as clinical and travel data helps the researchers to a great extent to predict the pattern of COVID-19 flow. There are also personal data that will be of help to analyze the affected status from social media channels such as their lifestyle, eating habits, family history and more.

From early detection of coronavirus to tracing the contacts, providing analytics on collected data from various sources, developing vaccines, AI plays an important role in fighting COVID-19 globally. Doctors and Governments are now using AI to trace the people who came in contact with the infected patients, to isolate and monitor them for symptoms to help stop the spread. There are various analytics techniques that are available to generate actionable insights to better study COVID-19 & fight against it. A few of them are described below:

  1. Path analysis – It helps to visualize the flow of coronavirus in infected patients. It projects in detail how the disease is progressing across a particular demography. Medical data of a region like daily reported cases, active cases, number of deaths & recovery are taken into account to do path analysis.
  2. Graph analysis – This analysis generates insights on the COVID-19 outbreak to predict the spread among the local population. A detailed analysis further identifies the super spreaders’ or events where the infection was super-spreading among the people. In the graph, analysis forms a network map that projects the degree of interconnectedness between infected cases & the contacts. Each junction represents an infected person wherein the edge represents the channeling of the infection through contact.
  3. Telco analytics – This refers to the analytics done with the data gathered from a person’s electronic devices such as mobiles, tabs, smartwatch & other electronic accessories. The electronic devices not only possess the personal data right from the name, age, gender, to GPS data collected both present & previous locations, to browsing patterns and much more. A lot can be analyzed with the GPS data including travel history & mobility patterns. The best thing about telco analytics is it is capable of providing real-time insights about an infected person’s activities who are referred to isolation or the suspects who are advised for home quarantine.

The applications of AI/analytics does not stop here. Researchers & therapeutics are betting on AI to discover potential COVID-19 medicines. AI is used to identify molecular compounds in the test data that can be as promising as to go for detailed testing in the labs within a few weeks of its discovery, unlike the conventional way where a team of scientists takes years of several experiments testing the chemical compounds in an effort to develop a new drug. Neural networks are used to accelerate the drug discovery process as they can run through the huge molecular data set & identify the required attributes to develop a new drug. NLP is being used to remove the texts from thousands & thousands of sources to cleanse news & statements made by representatives about the health of living beings.

Countries are broadly relying on AI applications to combat this infection because of their fast learning cycle, generating key insights & accuracy in forecasting from the collected data. AI is capable of learning both structured and unstructured datasets available anywhere in the sources. But one of the main challenges here is the availability of appropriate data by the government. A lot of manipulation happens in the data before it makes public in order to control the panic rising among the citizens. There are lots of open sources from where the data on coronavirus cases can be reaped using AI. Many AI-based firms are now developing a deep-learning-based model for automated detection COVID-19. The model is trained with chest CT scans images of COVID-19 infected patients from several hospitals treating COVID patients.

AI-powered chatbots are now widely being used to provide information about COVID-19 symptoms, do’s & don’ts during quarantine period & many more. Organizations like Facebook in association with the Indian government and national governments around the world are building chatbots to help people get accurate and timely information about COVID-19. They have also tied up with the World Health Organization (WHO) to get authoritative information about coronavirus directly in the WhatsApp inbox to help fight misinformation amid crisis. Real-time insights on daily reports about COVID-19 cases are being sent using this chatbot. Many healthcare tech industries are developing chatbots for people to go through self-assessment about COVID-19 symptoms. There are a set of parameters given if selected accurately the chatbot further guides the user if it requires to get admitted in hospital or home quarantine is applicable. The primary objective of a chatbot is to provide real-time information of COVID-19 cases, fight the fake information over social media handles & other sites, creating awareness & self-assessment of COVID symptoms.

Singapore hospital & public health facility are performing real-time temperature checks using smartphone & thermal sensors. China is using an AI system that uses sensor & AI to predict people’s temperature, this system can check the temperature of around 200 people per minute. The systems send an alarm if it detects a person with a temperature above 99.1 degrees Fahrenheit. Drones are being used to supervise people under home quarantine, thermal sensing & spraying disinfectant. As soon as South Korea’s government permitted private sector companies to begin developing coronavirus testing kits, a Seoul-based molecular biotech company used AI to accelerate the development of testing kits.

There is a race to find medicines and treatments not only in various AI tech companies & but also in between various countries like the US, China, South Korea, and Israel. Although every country in the world has adopted personalized approaches to end the disease, they are extensively using AI, Big Data & Data Science among various other technologies to combat COVID-19.

You can also help your country in this fight against coronavirus by regularly washing, sanitizing your hands, avoiding contact with any parts of your face and most importantly by maintaining social-distancing & avoid going out unless it is really necessary.

Leave a Reply

Your email address will not be published. Required fields are marked *