Healthcare professionals to get AI training as Singapore continues push towards preventive care

SINGAPORE: As Singapore moves towards preventive care for its people, artificial intelligence (AI) is set to help.

The country can harness AI to focus on delivering better diagnosis and treatment, early detection, prevention and prediction of diseases and faster drug development, said Dr Janil Puthucheary, Senior Minister of State for Health at the AI Health Summit on Thursday (Nov 23).

“The application of AI may take some more time to mature, and we are looking at how AI can help in the shorter term to enhance clinical decision support, increase productivity of our healthcare teams, and support individuals,” he said.

“Patients will be better and healthier residents in their Healthier SG journey for preventive population care.”

Ultimately, the impact of AI is felt through the healthcare professionals in delivering their work, he noted, adding that this means that they need to be equipped to make the best use of the new technology.

“They need to understand not only AI’s potential, but also its limitations and ethical implications of using these technologies,” he said.

To that end, under a Memorandum of Understanding (MOU) signed between AI Singapore — the national programme to boost AI capability — and SingHealth at the summit, healthcare professionals will receive AI training and qualification catered for their industry.

There will be support for joint research and innovation projects. The education programmes will start in June next year, and include a 10-part e-learning video, as well as physical classes and workshops. Topics include AI ethics.

USE OF AI IN HEALTHCARE

AI Singapore has developed training curriculum for the last couple of years, including certification and apprenticeship programmes, said its director for AI Innovation Laurence Liew.

“One of the reasons why we have this collaboration is really to bring our expertise on how to create the sort of rubrics and metrics to make sure that the AI professionals are properly trained within the regulations of the healthcare sector,” he told CNA’s Singapore Tonight.

Also joining the show, Associate Professor Daniel Ting, director of the AI Office at SingHealth, said that for the past 10 years, AI has shown great promises in helping healthcare professionals make better decisions.

For instance, AI algorithms can be run on mammograms to help screen for breast cancer, he told CNA’s Singapore Tonight. Retinal surgeons like himself are also able to detect changes in the eye among diabetes patients, he said.

Among the wider public, the use of smart watches and wearables is also helping in the push towards better and more preventive health, given it can monitor data like calorie intake and step count, he noted.

Technology like OpenAI’s ChatGPT could also potentially be applied to better communicate with patients, he said.

“We are currently developing some AI chatbots powered by large language models to help patients get … (their) questions answered pretty quickly,” he said.

CHALLENGES IN USING AI

When asked about the challenges of using AI in healthcare, Assoc Prof Ting said that knowledge on the biases in the technology is needed.

“We need to know how the AI was being built, because a lot of AI (models) come with biases. So how does each of the AI algorithms deal with the biases and whether this is a suitable or the right product for the right population is something that we actually are concerned about,” he said.

Data protection is another key factor in using AI, he said.

“Without data, you cannot build any AI algorithms. And for the data to flow into the AI system, we need to make sure that these data are being protected, wisely, and also safely,” he said.

Mr Liew said the curriculum will make doctors and nurses the allied health professionals aware of the safety and ethical concerns in the use of AI, and that users will also be able to give feedback when an AI model is not working well, or if it needs to be retrained.

When asked about the cost to patients from the use of such technology, Assoc Prof Ting said that while the upfront cost is “unavoidable”.

“But by preventing patients from going into hospital, by preventing the patients from going in, getting admitted to the ICU (intensive care unit), or preventing patients from developing any disease-related complications, the long-term costs, hopefully, we can actually justify, say in five, 10, 15 years’ time,” he said.