Taipei Veterans General Hospital (TVGH) in Taiwan is working to enhance outcomes for dialysis patients by using the NVIDIA Jetson AI model that helps in real-time heart failure risk prediction for kidney patients during dialysis. Taiwan has the highest prevalence of kidney dialysis patients based on density. To improve the procedure’s outcomes and provide better risk management for heart failure, TVGH hopes to mitigate cardiovascular risk as a leading cause of death in dialysis patients.
It plans to do so via an AI-based risk assessment model that achieves an accuracy of 90% and evaluates up to 200 sets of dynamic physiological and dialysis machine values while also processing medical records, blood test results, and medication data.
TVGH is adopting AI technologies like the NVIDIA Jetson AI Platform will enable TVGH to analyze data in real-time as it combines dialysis machine data with patients’ medical records and test results.
Read More: IIT-Mandi Announces MBA Program in Data Science and AI for the Upcoming Semester
Prof Der-Cherng Tarng, Chief of Department at TVGH, said, “By deploying NVIDIA Jetson next to each dialyzer to perform AI prediction during the procedure, we can achieve real-time insights in a way that’s affordable and effective, even for small-scale dialysis centers.”
While the AI model automatically records and analyzes data generated by the dialyzers, their initial workflow required the healthcare staff to note physiological changes every 30 minutes. To make the model provide real-time inference, TVGH adopted the Aetna Edge AI Starter featuring Jetson Xavier NX, which can process up to 21 trillion operations per second. TVGH’s team also used NVIDIA TensorRT software to optimize predictions for the platform.
The hospital is also working on more AI projects with NVIDIA Parabricks genomics software, NVIDIA FLARE for workflows, and the NeMo Megatron for NLP.
