A new realm of possibilities with AI in healthcare

Singapore's healthcare system, like those of all developed countries, is swamped with data. The challenge - and the opportunity - is to use that data to become a more effective premium healthcare provider. This can be achieved through the use of Artificial Intelligence (AI). AI lets medical professionals analyse and derive insights from the huge datasets created by healthcare organisations, a feat that is just not possible for humans.

AI analysis of clinical and operational data can result in the improvement of individual diagnosis, hospital administration, and even the production of drugs and other healthcare products. At a macro level, AI can be used to analyse demographic data, providing population-wide healthcare insights.

Singapore’s National University Healthcare System (NUHS) deploys multiple AI and automation tools as microservices on its ENDEAVOUR AI platform. These AI tools incorporate multi-domain patient information, such as demographics, text, images, lab data, and medications prescribed to provide a synthesis of a patient's condition. This translates into significant cost savings, from a patient’s care at admission, to predicting a patient’s length of stay, therefore optimising scarce bed resources.

 The application of AI has also helped companies like Pfizer revolutionise their supply chain to adapt and predict any winds of change. According to Hamid Mehdizadeh, Pfizer’s Director of Manufacturing Intelligence, “Pfizer researchers are able to shatter physical, chemical and biological limits to mass-produce life-preserving medication by making data-driven decisions through the digitisation and simplification of everyday tasks.”

Fighting COVID-19 with AI

The COVID-19 pandemic has proven to be an important test-bed for AI. Applications around the world have included dynamic, predictive heatmaps to help prevent the virus from spreading, and building predictive models of the citizens most vulnerable to the impact of the virus. Law enforcement has used AI to help enforce isolation measures, and AI has even been used to integrate weather reports with other demographic and medical data to predict a rise in cases due to higher ambient temperatures. 

In Singapore, AI and genomics helped the Ministry of Health to understand the spread of COVID-19 infections. The Ministry used AI to “identify the spikes” in respiratory infections, using natural language processing to analyse the notes made by staff at emergency clinics. The AI could identify symptoms of respiratory conditions by picking out key phrases in these notes.

Another Singapore example is VigilantGantry, developed by GovTech. This is an AI-driven automated temperature screening gantry that augments existing thermal systems to enhance the rate of contactless screening, saving time and manpower. The solution can interface with facial recognition software to enable contact tracing.

As we move into the endemic phase, the use of AI and machine learning (ML) solutions can be used to check how ready employees are to return to the office. Employees with symptoms can provide daily updates and get advice on whether they can come back to the office or must continue working from home.

Keeping the cost of healthcare in check

The application of AI in healthcare does more than just provide added health benefits. The efficiency of AI is also required for Singapore to rein in the rising cost of medical care and treatments.

Singapore's Health Minister, Ong Ye Kung has said that the national healthcare expenditure has more than doubled from SGD 10 billion in 2010 to SGD 21 billion in 2018 and is expected to triple to SGD 59 billion by 2030. This exponential rise needs to be addressed, and AI can play a significant part in helping drive cost-saving efficiencies through increased diagnosis accuracy, earlier detection of diseases, boosting of life-saving decision making and the reduction of repetitive day to day tasks.

 A three-year collaboration between SingHealth and SGInnovate will offer resources and opportunities that deep tech startups need, to develop AI applications to enhance healthcare services. The partnership will drive the adoption of AI and other emerging technologies to improve diagnostics and treatment, as well as healthcare delivery and clinical outcomes for Singapore, said Health Minister Ong Ye Kung at the recent signing of the Memorandum of Understanding.

In 2021, the government announced that it will invest SGD 180 million (USD 133.31 million) in the national research and innovation strategy to tap the technology in key areas, such as healthcare and education.

AI’s role in healthcare is ever-changing, and medical practitioners need insights from data to make life-saving decisions. Making these insights widely available is essential for people, processes and systems to act without delay. The rewards of overcoming this challenge are transformative. AI can do much of the heavy lifting for healthcare professionals – identifying gaps and freeing up resources, with the result of a better focus on patients and improved outcomes.

Regulatory hurdles for healthcare AI

All facets of the healthcare industry are highly regulated, and it will be the same with healthcare AI. Technological capabilities are just the start of actualising AI's benefits - the same as every technology used in healthcare.

Singapore's Ministry of Health, Health Sciences Authority (HSA), and the Integrated Health Information Systems (IHiS) have co-developed the MOH Artificial Intelligence in Healthcare Guidelines (AIHGle) (read as ‘agile’). The AIHGle aims to support patient safety and improve trust in the use of AI by sharing good practices with Developers (e.g. AI Medical Device manufacturers) and Implementers (e.g. Healthcare Institutions), complementing HSA’s regulations for AI Medical Devices and periodically being updated as a ‘living’ document.

Still, there are potential biases in AI programming. It’s important that algorithms are seen to be free from bias. And as AI in healthcare will involve sensitive and personal data, it is essential that best practices in data governance and security be put in place. As enterprises deploy and integrate AI into everyday activities, it is crucial that accountability and best practice be maintained to ensure public trust as businesses actualise the vast potential of their AI programmes.

As patients become more comfortable with digital services for complicated and delicate matters, AI’s future potential in healthcare can only grow brighter.