News
News Categories

NUHS's new NVIDIA-based AI production platform improves patient care and treatment

By Ken Wong - on 30 Nov 2021, 1:19pm

NUHS's new NVIDIA-based AI production platform improves patient care and treatment

Singapore’s National University Health System (NUHS) has built an AI production platform based on NVIDIA’s DGX A100 system.

As the heart of NUHS’ newly-launched Endeavour AI platform, NVIDIA DGX A100 will run the AI tools that make real-time predictions on diagnosis, progression of diseases, readmissions, risk of falls, and others.

This allows NUHS to be the first healthcare group in Singapore to have real-time streaming capabilities to deliver better patient care and treatment, collaborate on biomedical research and transform how illnesses are managed and treated.

NUHS is one of three public healthcare clusters in Singapore and an integrated academic health system and regional health system covering 19 hospitals, polyclinics, specialist centres, medical centres, and academic health science institutions.

NUHS says that Endeavour AI will be integrated with NUHS’s Discovery AI training platform to form a complete training and inference system as part of the group’s digital transformation. Discovery AI, which runs on NVIDIA GPUs, is used for training models using large data sets while CPUs in another system were used for inferencing.

Now, Discovery AI will continue to be used for training with Endeavour AI taking over the inferencing.  

 

Powering Endeavour AI

Image source: NVIDIA.

Each NVIDIA DGX A100 in Endeavour AI is a 6U form factor packing five petaflops of AI performance doing everything from analytics to training to inference. This allows it to replace legacy infrastructure silos with one platform for every AI workload.

In total, Endeavour AI is a software and hardware stack that features streaming data as well as AI tools running microservices off a Kubernetes backbone to process all the streaming data and produce outputs on a real-time basis.

According to NUHS one AI tool would run about 100 to 200 inferences per second. For every patient who turns up at its hospitals and polytechnics, every time a doctor clicks, saves or free texts, or when new lab test results are out, an AI tool will be running in the background. All the data gets processed by AI tools. This is done hundreds of times per second throughout the whole cluster at a large volume. If only CPUs are used, NUHS will run out of processing speed very quickly.

Dr Ngiam Kee Yuan, group chief technology officer of NUHS and deputy chief medical informatics officer of National University Hospital (NUH). Image source: NVIDIA.

Dr Ngiam Kee Yuan, group chief technology officer of NUHS and deputy chief medical informatics officer of National University Hospital (NUH) said:

As the number, volume, and speed at which we are running inferencing increase, GPUs become necessary. Otherwise, we need to expend a lot more CPUs to run the same inferencing at that speed. That's why we strategised, planned and built-in NVIDIA DGX A100 from Day One when we deployed Endeavour AI. This is because we will be using it for high speed and large volume inference processed by our AI tools.

Since the AI tools went live there have been a number of expected benefits and improvements said Ngiam.

AI-powered chatbots interaction was improved with faster appointment making and reduced waiting times. Day-to-day hospital operations saw predictions being done automatically without even needing to click a button.

With real-time predictions on diagnosis, progression of diseases, readmissions, and risk of falls, clinicians and radiologists can make faster and better decisions, resulting in improved patient care and treatment.

Join HWZ's Telegram channel here and catch all the latest tech news!
Our articles may contain affiliate links. If you buy through these links, we may earn a small commission.