The global COVID-19 pandemic has gripped the world by fear and forced many individuals and businesses worldwide to rethink what their new normal is. But one thing’s for sure: healthcare systems are cracking under the pressure. Hospitals are now greatly understaffed, overwhelmed with patients, and critical equipment is in short supply.
Over the past few months, the coronavirus has already infected more than 2 million worldwide, and caused at least 200,000 deaths. As the figures continue to rise and greater uncertainty looms over the global economy, the fight against the virus is now at the forefront of worldwide efforts. From healthcare workers on the frontline to medical researchers on the quest for a vaccine, the crisis has opened up a stage for international cooperation like never before.
Yet amid this crisis, one player emerges as a beacon of hope in this ongoing challenge: Artificial Intelligence (AI).
How is AI making waves in healthcare now?
AI’s ability to handle both structured data and unstructured information through its Machine learning (ML) and Natural language processing (NLP) methods opens up a whole world of possibilities in the field of healthcare.
In recent years, a burgeoning number of corporations have developed AI tools that are capable of processing huge quantities of information and delivering analysis at break-neck speed. These tools are capable of examining virtually anything, from quicker and more accurate ways to diagnose patients, to new findings in the mission to cure cancer.
In fact, it was AI that first identified the emerging cluster of infections in China. Bluedot, a Canadian-based AI platform, predicted the outbreak and sent an alert to its clients on December 31, several days before the World Health Organisation (WHO) notified the public on January 9.
Even during the ongoing pandemic, AI’s potential is being harnessed by governments, medical professionals, researchers, and companies worldwide to help prevent the spread of COVID-19, accelerate the discovery of a cure, and aid in decision-making.
AI can help limit the spread of viruses
AI has been, and may further be used, to contain the pandemic through faster identification and diagnosis, as well as minimising human interaction on the frontline.
Cases screening & diagnoses at scale
When the Wuhan Municipal Government ordered for all suspected cases to be screened by gathering information of residents’ symptoms through phone calls, operators were met with an overwhelming task that they could not complete – making hundreds of calls in a day. In addition, the quality and efficacy of these calls would differ, depending on the judgement and experience of the operator.
To counter this, Ping An Smart Healthcare developed a smart audio screen system for COVID-19, boosting the ability to prevent and control the outbreak. Built on AI technology, the system will trigger automated dialogue based on programmed questions, interpret the conversation’s semantic context, format and categorise the findings for risk evaluation, and report to the management team who can then monitor residents with reported symptoms accurately.
Chatbots present another way for workers to lessen their workload and improve productivity when screening for patients with virus related symptoms. This can be especially useful for infected patients who face a reduced ability to convey information vocally due to symptoms such as shortness of breath, as well as facilitate home care with efficiency for countries where cases are distributed outside of urban cities. With the use of semantic technology in these chatbots, they are poised to respond to a wide range of enquiries with humanistic accuracy. For example, NLP-based chatbots, like TAIGER Converse, can comprehend complex sentences while taking into account unique contextual information.
Apart from more effective screening within the population, recent research from the UN Global Pulse has shown that AI can also help with diagnosis, offering the same precision as humans while reducing time and expense as with regular COVID-19 tests. In that sense, AI has doubled up as a doctor’s assistant, providing an extra ‘brain’ to help sieve through results.
Chinese tech giant, Alibaba, has developed an AI tool capable of detecting the coronavirus in chest CT scans, with the system already being implemented in at least 100 Chinese hospitals. Researchers who created the program claim that the AI has a 96% accuracy.
Another AI system, COVID-Net, was built to detect COVID-19 cases from chest x-ray images, and utilised data from patients with various lung disorders, including COVID-19.
Localised to city-wide containment management systems
On the frontline, AI is helping to minimise human interaction that could further aggravate the spread of the coronavirus.
“Social distancing detection” tech is now available thanks to Camio, a US-based computer vision startup. The software uses camera images to detect when social distancing norms are being violated, and thereafter sends an alert out.
Elsewhere In China, robots roam the streets testing residents for fever and other COVID-19 related symptoms, and offering free hand sanitiser. They are also being used by Chinese companies to perform contactless delivery and to disinfect public spaces, reducing the chances of cross-infection.
In hospitals, robots are taking the strain off medical workers and reducing their chances of being infected. One such example is the Alexandra Hospital in Singapore, which has enlisted a robot called BeamPro to deliver food and medicine to infected patients, or those who are deemed as suspected cases. The robot even enables patients to have “face to face” interaction with doctors and nurses via the screen and camera, redefining what it means to be physically isolated yet socially connected.
AI can help speed up the process of finding a cure
Long before the COVID-19 outbreak, AI was lauded for its potential to contribute to new drug discovery. Now, researchers around the globe are racing to find effective treatments for the novel coronavirus, with laboratories and data centres tapping on AI to obtain breakthroughs for a vaccine.
AI’s potential in the quest for treatment solutions is perhaps most apparent in the example of Google’s AI research lab Deepmind, which predicted COVID-19 protein structures through its state-of-the art system, AlphaFold. This knowledge can be extremely useful in the development of new drugs to fight the virus.
Another firm, Singapore-based AI platform Gero, has identified a variety of potential drugs that can be used to treat COVID-19, such as Afatinib, a medication commonly used in the treatment of lung cancer.
To further aid researchers worldwide in the mission for a cure, the White House Office of Science and Technology has set up the COVID-19 Open Research Dataset (CORD), a freely accessible data repository composed of over 24,000 papers and is constantly being updated. As this data set is machine-readable, researchers can build and implement NLP algorithms to consolidate, retrieve, and analyse key findings from reports.
“Something like this crisis shows how AI can potentially do a world of good,” said Oren Etzioni, CEO of AI2, in an article by Wired on how various organisations are leveraging AI and machine learning to evaluate thousands of COVID-19 research papers.
Retrieving information from such a diverse dataset can be made easier through the use of a semantic search engine such as TAIGER Search. Going beyond mere keywords, this tool understands complex queries, relates contents, expands searches and recovers corresponding documents more efficiently and precisely.
Findings from research reports can also be extracted and consolidated through intelligent information extraction tools that automatically identify, extract, cleanse, validate and store key pieces of knowledge required with a high accuracy rate. This reduces the total processing time needed and helps researchers dive quicker into their work. See how TAIGER Extract reduced time spent on manual extraction by 98%.
AI can help decision-makers make better and faster decisions
As the number of infections continues to climb, real-time data becomes a valuable source of information for governmental bodies and medical professionals. By shortening the time critical data is made available, experts are able to study the dynamics of the disease, identify patterns, and ultimately anticipate societal impacts more promptly. Consequently, decision-makers are able to make informed decisions rapidly, enhancing response efforts towards the pandemic.
Many organisations have made it easier for us to view at a glance the COVID-19 situation and its related numbers through data dashboards that help people better visualise the pandemic. Some notable ones as ranked by MIT Technology Review include that of Upcode, NextStrain, and the John Hopkins Center for Systems Science and Engineering.
“We can run simulations unlike we’ve ever done before, we understand biological pathways unlike we’ve ever understood before, and that’s all because of the power of AI,” said Michael Greeley of the venture capital firm Flare Capital Partners, which has invested in a variety of medical AI startups.
The future of AI in healthcare
The healthcare industry has already been transformed by the many applications of AI, and will continue to with the advancement of its technologies. Based on a report by Frost & Sullivan, the market potential for AI in healthcare is huge and is projected to reach $6.6 billion by 2021.
Even in a post-pandemic world, AI will remain pivotal in helping humans enhance processes and solve problems. In an interview with The New York Times, Eric Topol, Founder of Scripps Research Translational institute, even suggests that AI can bring back the humanistic aspect that has eluded healthcare professionals for some time.
“By augmenting human performance, AI has the potential to markedly improve productivity, efficiency, workflow, accuracy and speed, both for doctors and for patients.”
The pandemic has indeed laid bare the cracks present in healthcare systems around the world. It is alarming to watch medical workers bear the brunt of the COVID-19 pandemic and work doubly hard to make up for the weaknesses of the system. This presents an urgent call for leaders to reevaluate current medical infrastructures, and reinvent their systems to be built-in with insight and foresight. Given AI’s remarkable potential, healthcare systems leveraging it can be transformed to better meet evolving needs in the near future.