Physical Address

304 North Cardinal St.
Dorchester Center, MA 02124

National Taiwan University Hospital to go multimodal in large AI development


National Taiwan University Hospital is moving to the next phase of its large AI development to focus on multiple data types. 

The hospital recently acquired two new supercomputers from NVIDIA, which would help accelerate its development of smart healthcare applications. 

This includes projects developing multimodal large language models for optimising operational processes and enhancing patient service quality, NTUH shared in a statement.

THE LARGER TREND

NTUH first acquired NVIDIA’s AI supercomputers in 2020, which led to its development of various generative AI and extended reality-based applications. 

Its Center of Intelligent Healthcare, for example, has developed LLMs for ICD-10 automatic coding, automated health check-up reports generation, auto-generated telemedicine consultation transcriptions, emergency voice record reporting, and extraction of key points in pathology reports. NTUH’s IT Office has also applied LLMs to summarise medical records, generate reports, mine unstructured data, and answer medical questions. These LLMs have been integrated into NTUH’s Health Information System. 

With its supercomputing capability, the hospital has also developed a VR platform for surgical training called OpVerse. 

Meanwhile, two top electronics manufacturers in Taiwan recently announced projects to ambitiously build what could be the country’s largest supercomputing capability. Foxconn, partnering with NVIDIA, aims to develop the fastest AI supercomputer in Taiwan with over 90 exaflops of expected AI performance, potentially supporting cancer research and LLM development. Just last month, Foxlink also unveiled its first supercomputing facility, Ubilink – also powered by NVIDIA – claiming to have Taiwan’s highest computing power to date.

Outside Taiwan, two major healthcare clusters in Singapore, SingHealth and the National University Health System, have utilised supercomputers in recent years to speed up their development of AI applications, including a genAI-based chatbot and digital twin technology for monitoring disease outbreaks.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *