Achieves a 79,000x reduction in energy per inference for insulin dose calculation using Spiking Neural Networks (SNNs).
March 31, 2026
Original Paper
An Energy-Efficient Spiking Neural Network Architecture for Predictive Insulin Delivery
arXiv · 2603.27589
The Takeaway
While achieving slightly lower accuracy than LSTMs, the massive energy savings (femtojoules vs nanojoules) make continuous, on-device medical monitoring viable for wearables. It demonstrates the practical trade-off where neuromorphic architectures enable capabilities that are otherwise power-prohibitive.
From the abstract
Diabetes mellitus affects over 537 million adults worldwide. Insulin-dependent patients require continuous glucose monitoring and precise dose calculation while operating under strict power budgets on wearable devices. This paper presents PDDS - an in-silico, software-complete research prototype of an event-driven computational pipeline for predictive insulin dose calculation. Motivated by neuromorphic computing principles for ultra-low-power wearable edge devices, the core contribution is a thr