Coronavirus: How AI and Data Science can help to fight the Pandemic
- March 23, 2020
- Posted by: admin
- Category: Machine Learning
A deadly “CORONAVIRUS” on December 2019 appeared for the first time in Wuhan, China.
This corona virus disease (COVID-19) is rapidly spreading to others parts of the world at an alarming rate.
With the help of Artificial Intelligence: Data science, Machine Learning and Deep learning we can definitely wrangle data and get more insights about the disease that can speed up the process to uproot this problem which has hampered our life.
Data is fuel to any machine learning model so it’s very important to understand all the aspects before diving deep into the technical details.
Few facts about Coronavirus:
The people who are infected by this disease suffers from respiratory illness.
According to WHO (world health organization) till date – 21 march 2020 there are 2,67,013 cases registered out of which confirmed deaths are 11,201 worldwide.
It has reached more than 100 countries but the number of cases has surged in the Asia and Europe.
The top ten countries are – China, Italy, Iran, Spain, Germany, USA, France, South Korea, Switzerland, UK. These all countries have highest number of COVID-19 cases.
The common symptoms of this disease are Fever, Fatigue and Dry cough. Some patients also face diarrhoea, sore throat, runny nose, shortness of breath, aches and pains. The virus is thought to spread when there is person to person contact – The person who is infected coughs or sneezes and through the droplets the virus can get transmitted to other person when they land on hand or mouth of a person.
The virus also spreads when person comes in contact with the contaminated objects or surfaces. As it lives on the surface that’s why its so contagious.
According to study it stays on plastic for 72 hours, on stainless steel surface and cardboard for 48 hours. As it is highly contagious the people are advised to avoid public places, use masks to protect, sanitize public places, wash your hands frequently, don’t share your stuffs, avoid touching your face and social distancing.
As around the globe coronavirus cases are increasing day by day in addition to the above precautionary measures the people are advised self – quarantining to curb the spread of COVID-19.
Especially the persons who have travelled to other countries and those who came in contact with sick person must self – quarantine.
The person over the 60 years of age, infants and person with medical conditions are at higher risks. The incubation period of coronavirus is between 2 – 14 days and average incubation period is 5 days.
These are facts and figures of coronavirus. Apart from this data lots of data is generated everyday on the media platform about the virus. If we utilize all this data then we can definitely curb the outbreak.
Before anyone else knew about it the BlueDot start-up spotted the coronavirus. The Canadian start-up BlueDot tracks infectious disease by using AI service.
An algorithm developed BlueDot predicted the outbreak of coronavirus and sent an alert to client on 31st December.
BlueDot is a Canadian start-up that tracks the infectious diseases around the world. It takes information from thousands of sources like health organizations, flights, climatic conditions, health reports etc. and tracks infectious diseases around the world by using natural language processing and machine learning.
It learns the patterns from the data and predicts and forecasts as well. By studying the travel history and learning the patterns AI predicts the next travel trip of a person who is contracted with coronavirus and in this way predicts the spread of virus across globe.
Facial recognition system and temperature detection software is used in conjunction at china surveillance system to track coronavirus. Computer vision technology and sensors are used in this project.
The Chinese government is accessing the travel history of an individual to identify potential threat. The monitoring system accesses the travel pattern of an individual using big data and by checking how much time the person was closed to infected or virus prone areas. The citizens are then assigned a health colour code (red, yellow and green) which indicate whether they should quarantine or not.
Thermometer guns are used for the screening of coronavirus. Baidu has developed AI system with sensors and computer vision technology to check the temperature of a person. 200 persons can be checked within one minute with the help of this AI system.
Another AI system for detection of coronavirus in chest CT scans is developed by another Chinese tech gaint Alibaba. The system is trained on 5000 cases of corona virus patients and perform the test within 20 seconds. With 95% accuracy another Chest CT scan AI system is developed by researchers from Renmin Hospital of Wuhan University, Wuhan EndoAngel Medical Technology Company, and the China University of Geosciences.
Deep Mind Recently declared that it has learned about the structure of protein associated with COVID-19 using deep learning. It is vital information and can thus help researchers for preparing vaccine formula.
Different coronavirus datasets are available and with the help of that we can analyse the outbreak across various regions, using visualization techniques we can create pictorial representation, predict the number of upcoming cases using machine learning algorithm.
Insilico Medicine uses AI for drug discovery. Drug Development takes lots of effort and time to move the idea to market.
As well as there are so many factors which decides the success rate of this experiment. “We can substantially accelerate this process using AI and make it much cheaper, faster, and more likely to succeed,” says CEO of Insilico Medicine, Alex Zhavoronkov. This company published a paper in just 4 days related to molecules that could fight coronavirus. AI based system at Insilco took four days to come up with a conclusion and to identify thousands of new molecules that could be turned into potential medicine against COVID-19.
Similarly, companies who are persistently putting efforts for the discovery of coronavirus drugs are:
Deargen is using deep learning models to develop new antivirals. SRI International and AI company Iktos in Paris announced a collaboration to discover and develop new anti-viral therapies on coronavirus.
Graph analysis can be done in understanding and forecasting the spread of virus across different geographical location during outbreak. Network map is developed by Singapore coding academy where each node in a graph is a patient and edges represent the path of infection through contact.
Robots are used at Chinese hospital to help combat outbreak.These robots are used to protect medical staff from coronavirus infection. These robots are used to deliver medicines, food to the patients , disinfect rooms and for communication purpose with isolated patients.
UVD robots are also used to disinfect rooms by emitting ultraviolet rays. Along with this drones are also used to sanitize places, thermal imaging and patrolling. Chinese officials are using drones to check if someone is roaming without masks or if someone is violating quarantine rules. Different technologies like computer vision, sensors, deep learning models and used in conjunction to develop such types of machine.
AI, Data science and Machine Learning can be used in different stages of the outbreak.
It can predict the count of patients in the upcoming days, forecasts, helps in drugs discovery, analyse history of smartphones and dig information about travel trips, maps relation between different points of data and can thus help to eradicate the disease. Thus, it helps in all the phases right from prediction to recovery.