NUHS to leverage edge supercomputer to handle AI workloads

Clinical researchers at Singapore’s National University Health System (NUHS) will soon be able to harness cutting-edge supercomputers to accelerate the formation of artificial intelligence (AI) models in fields such as radiology.

When ready by the middle of next year, the state-of-the-art petabyte-scale supercomputer infrastructure, dubbed Prescience, will reside on NUHS premises to meet data privacy and security requirements. .

Speaking at a press conference on the collaboration between NUHS and Singapore’s National Supercomputing Center (NSCC) to build Prescience, Ngiam Kee Yuan, group chief technology officer at NUHS, said supercomputing capabilities advanced can be used to predict patient health trajectories, healthcare robotics as well as the calling of genomic variants.

Additionally, Ngiam said the NUHS will use Prescience to train conversational chatbots that are able to converse with people in a more natural way. “These conversational chatbots have many use cases – for example, advising patients on chronic disease care,” he said.

Prescience will be powered by graphics processing units (GPUs), which are faster and more efficient than central processing units to handle AI workloads in parallel.

At NUHS, training AI models with large amounts of data typically takes days, but Ngiam said the high-performance computing infrastructure will cut training times to hours, allowing his medical and paramedical staff to ” optimize patient trajectories and improve patient care.

Ngiam said Prescience will also enable NUHS to undertake large-scale AI projects, making the anonymization of data used to train AI models more effective.

The NUHS uses a platform called Discovery AI to anonymize patient data, but this creates overhead in training AI models.

“If I want to send my data to the central supercomputer, I have to do a large amount of data pre-processing first before I can publish it,” Ngiam said, noting that having supercomputing capabilities onsite allows data to be processed natively at its source.

Bernard Tan, NSCC director of strategy, planning and engagement, said technical details of the edge supercomputing infrastructure were being finalized with the NUHS and more information would be shared later. .

In addition, the NUHS is also working with Singtel to deploy an indoor 5G network with advanced multiple-access computing capabilities in operating rooms and wards at the National University Hospital.

As a first for a public hospital, the rollout of 5G would resolve latency and bandwidth limitations and open up possibilities for better healthcare experiences, such as smoother teleconsultations, increased surgical navigation at the hospital. using mixed reality devices and robotic AI capabilities using cloud and advanced computing.

Comments are closed.