2020 has been a year of innovation for Singapore’s healthtech agency, IHiS. It helmed the development of many tools to help frontliners combat the virus and contain its spread.
From setting up community care facilities in five days to building AI temperature scanning devices, tech has been the bedrock of its response. The agency has made significant “headways since the start of its healthcare AI and analytics journey” a few years back, says Andy Ta, the agency’s Director of Data Analytics & AI.
GovInsider caught up with Ta to find out what’s next in Singapore’s AI and data journey in healthcare, and how it will tackle bias and privacy.
While Covid-19 has disrupted life as we know it, it has presented “unique conditions” for Ta’s team to come up with AI and analytics solutions in healthcare.
IHiS worked with Tan Tock Seng Hospital to build the Command, Control & Communications (C3) system used in the Operations Command Centre. C3 supports the National Centre for Infectious Diseases, which is responsible for a large part of Singapore’s Covid response.
With the help of AI and analytics, C3 functions like “the brain of the hospital”, coordinating day-to-day activities to free staff up from making multiple phone calls. That way, they can focus on handling complex problems in patient care, Ta says.
The hospital has used analytics and AI to predict the discharge of Covid-19 patients, and the daily and hourly arrivals of non-Covid patients. That has helped to predict bottlenecks before they occur.
At the peak of Singapore’s Covid-19 outbreak, it built community care facilities to house infected patients. Staff at the Expo facility have used analytics to improve the visibility of patients’ vital signs through alerts, Ta says.
That has allowed the limited number of healthcare personnel onsite to effectively monitor and focus on patients whose condition is at risk of deteriorating, he adds.
IHiS has developed OK Lah!, a risk score profile for visitors based on their travel and contact history. This risk scoring is incorporated into an existing travel declaration form used by public healthcare institutions, he shares.
Other healthtech innovations from IHiS include AI that predicts pneumonia severity; a central Covid-19 test repository to help with operations planning; and an interactive dashboard that tracks critical medical supplies in hospitals. The dashboard was created in partnership with public healthcare supply chain agency ALPS.
IHiS’ work on C3, the Covid-19 test repository, and its AI temperature scanning tool iThermo, were recognised with three awards from the recent Singapore Computer Society’s IT Leader Awards.
Tackling privacy and bias concerns
Now that the AI and analytics healthcare ecosystem has matured, IHiS plans to focus its efforts on tackling privacy and bias concerns.
AI learns to make decisions based on data, which may include biased human choices or social inequalities. Amazon had to scrap its AI recruiting tool in 2018 after it discovered a bias against women, Reuters reported.
If data is not anonymised properly, AI can also magnify the risk of privacy violations.
The “dynamic and diversified nature of healthcare” can make it complicated to implement analytics and AI at times, Ta notes. But ensuring medical confidentiality is crucial for maintaining patient trust.
IHiS has built an analytics platform that enables analysts to request for information from disparate data sources, Ta says. Known as the Business Research Analytics Insights Network, it anonymises data prior to further analysis, ensuring confidential data is kept private.
Before every AI or analytics solution is deployed, IHiS will verify the model’s result against real-world outcomes for consistency and reliability. It will conduct pilots to uncover potential teething issues; and a post-deployment review will ensure the intended outcomes have been achieved, Ta says.
AI ethics is a “crucial knowledge area” that Ta’s team intends to build up more in 2021. That may help to build trust and confidence to adopt AI, he notes.
IHiS also plans to work on explainable AI to ensure that algorithms make fair and transparent choices, said Christine Ang, Deputy Director of IHiS’ Emerging Capabilities-Health Insights. This is especially important in healthcare, where doctors would need to explain decisions to patients.
Genomic and medical imaging
Ta’s focus in 2021 is to continue finding areas in healthcare where AI will make the most impact. His team plans to expand into the medical imaging and genomics space, as they “foresee a lot of opportunities” in using analytics and AI to analyse genomics data.
For instance, genomics can help determine the risk of Down syndrome in pregnancy, and predict one’s likelihood of suffering from a disease, public sector R&D agency A*STAR told GovInsider.
Collaborations for a new frontier
As the national healthtech agency, IHiS will be the “core pillar” to accelerate the use of AI in healthcare. Collaborations will be crucial in this effort, Ta says.
IHiS issued a call for collaboration late last year to invite interested companies to develop an AI-enabled medical imaging platform.
The agency also plans to work closely with public health partners to enhance Singapore’s AI and analytics capabilities. Its Multiple Readmissions Predictive Model, for instance, has been used to identify patients at risk of readmission.
This enables earlier interventions – patients can be placed into community care programmes upon discharge. Nurses can also visit them at their homes to educate them on how to care for themselves or arrange for further social support.
Covid-19 has underscored the need for an innovative, responsive healthcare system. With IHiS’ ambitions to advance the use of analytics and AI, the nation will be better equipped to respond to the challenges of tomorrow.