A crucible of ideas
In the fight against Covid-19, Artificial Intelligence is emerging as a dominant tool across the globe. The technology is usually used to train computers to understand pattern recognition, interpretation, and prediction using Machine Learning, Natural Language Processing (NLP) and Computer Vision. AI application are helpful as they help in effective diagnosis and provide medicos sufficient data to treat Covid-19. AI can also assist in managing socio-economic impacts too in a pandemic situation.
Since the outbreak, there has been a rush to explore and unleash AI enabled tools to cure and address the soaring demand of patient treatment infected by novel coronavirus (nCov-2). Medical AI is not a new concept and has been around for long. The current trend already shows how AI has the ability to quickly adapt to cutting-edge technology.
Chinese technology giant called ‘Baidu’ has developed an AI driven sensor, which is a no-contact infrared sensor system, which can quickly recognize individuals with a fever, even in a crowd. In the same way, Florida based, Tampa General Hospital, deployed an AI system, installed at entrances to capture individuals with potential symptoms of Covid-19. By positioning cameras at entrances, the technology performs a facial thermal scan and differentiates other symptoms, such as sweat and discoloration, to restrict visitors who show signs of fever.
AI-based systems, which can analyze CT images within 20 seconds, with accuracy rate of above 90%, has now come to the rescue of doctors and radiologists for speedy diagnosis. Usually doctors can identify the disease from CT scan images from features like a shadow over a patient’s lungs. A single patient can have about 300 CT images, which can take a doctor a lot of time to analyze with the naked eye. For COVID-19 patients, radiologists also need to check earlier scans, putting more pressure on diagnosis. Here AI based systems recognise these patterns in the images in a jiffy, thus speeding up diagnostic processes without compromising on accuracy.
From early on, the Chinese government has been encouraging development of AI tools for healthcare. Its not the Chinese alone, but tech giants and healthcare companies in various countries have begun augmenting their efforts to develop AI systems to accelerate processing of CT scans of suspected COVID-19 cases.
At Zhongnan Hospital of Wuhan University in Wuhan, the radiology department uses AI software to detect early visual signs of pneumonia, usually associated with Covid-19 by studying images from lung CT scans. According to Haibo Xu, professor and chair of radiology at the Hospital, the software helps screen patients and prioritize those who are likely to be suffering from Covid-19. Detecting pneumonia on a scan alone doesn’t confirm that a person has the disease, but Xu says that doing so helps staff diagnose, isolate, and treat patients more quickly.
An AI-based tool, developed by three bio-tech students and a professor from Mumbai, can test COVID-19 through voice-based diagnosis using a smartphone. This was reported by PTI on April 16 . The patented technology is now being used by University of Tor Vergata in Rome, and has already been tested on 300 individuals there. The tool was developed by students from DY Patil Institute of Bio Technology and Bio Informatics.
Recently a group of scientists including one of Indian origin in Toronto in Canada collaborated to use artificial intelligence (AI) to identify an underlying genomic signature for 29 different DNA sequences of novel coronavirus that causes the disease. This was an important data set for vaccine and drug developers. The new AI tool will allow researchers to speedily classify viruses like SARS-CoV-2 in just a few minutes, according to researchers, including Gurjit Randhawa from Western University in Canada. Randhawa’s contribution was recently reported by ndtv.com on April 29, 2020.
Recently a Gurugram based startup Staqu, launched a unique thermal camera that uses AI to track potential suspects. The technology, reported by www.livemint.com, uses video analytics for a risk-free identification of Covid-19 suspects and avoids human contact.
Given the urgency, tech giant Microsoft is mobilizing it’s ‘AI for Health’ initiative to focus on helping those on the frontlines of research of COVID-19. “We’re focusing our efforts in five specific areas where we think data, analysis and the skills of our data scientists can have the biggest impact. And we’re immediately dedicating $20 million to this specific effort,” claims a statement from the tech company.
The Republic of Korea has largely been successful in containing COVID-19 without shutting down its economy. The country’s experience from the previous SARS and MERS outbreaks came handy during the current pandemic, and it also used technologies like AI efficiently. An important part of the Korea’s strategy was widespread testing. According to life sciences company Seegene, which came up with testing kits in under just three weeks, developing such kits would not have been possible without AI.
Special partnerships :
• The COVID-19 High Performance Computing Consortium: Its a public-private effort spearheaded by the White House Office of Science and Technology Policy, for which Microsoft is providing researchers access to the world’s most powerful computing resources, that can significantly speed the pace of scientific discovery in the fight to stop the virus. Around the world, Microsoft’s research scientists, spanning computer science, biology, medicine and public health, are collaborating on projects in the consortium
• The Institute for Health Metrics and Evaluation (IHME): A global health research organization at University of Washington School of Medicine, is continuously releasing COVID-19 data visualization sets and forecasts that the White House, FEMA, governors and hospital administrators have started using to mobilize resources
• The Sepsis Center of Research Excellence (SCORE-UW): The effort is part of University of Washington’s Department of Medicine, and involves a global collaboration between a network of hospitals, industry, blood banks, universities and funding partners. The effort uses clinical data, radiologic imaging and other patient biomarker responses. SCORE-UW develops algorithms to predict, and improve, healthcare and socioeconomic outcomes of COVID-19 positive patients