Efficient 3D imaging on low-power edge devices speeds up CT scans
An Efficient Out-of-Core Tomographic Imaging Framework for Edge Devices
Distributed, Parallel, and Cluster Computing
Summary
Computed Tomography (CT) scans create 3D images used in medicine and research but usually require powerful computers. The authors developed a new method called edgeFBP that works well on small, low-power devices like Nvidia Jetson, which have limited memory and processing power. Their approach speeds up the image reconstruction process using special computing units and smart precision choices, making it much faster and more energy efficient compared to current solutions. This allows advanced CT imaging to be done in settings without big data centers.
What this means in practice
- •For medical device developers: Integrate efficient CT image reconstruction on low-power embedded devices for portable medical scanners in remote or resource-limited areas.$Commercial implications: Enables development of portable CT scanners with lower power consumption, making them sellable to healthcare providers in low-resource settings.
- •For industrial inspection teams: Deploy compact CT imaging for non-destructive testing on edge devices directly at manufacturing sites without needing large computing infrastructure.
Authors
Xuetao Chen, Cong Ma, Xiangyu Meng, Du Wu, Zhengyang Bai, Tao Luo, Zhaorui Zhang, Emmanuel Jeannot, Edgar Josafat Martinez Noriega, Xun Wang, Peng Chen, Amelie Chi Zhou, Mohamed Wahib
Abstract
Computed Tomography (CT) is an essential 3D imaging technology widely used in medical diagnostics and scientific research. However, performing CT imaging on edge devices is challenging due to limitations in computational power, memory capacity, and energy budget. This paper presents an efficient CT reconstruction framework, called edgeFBP, designed for Nvidia Jetson System-on-Chip (SoC) devices. edgeFBP adopts an end-to-end pipeline design for efficient out-of-core image reconstruction under tight power and memory constraints. edgeFBP utilizes a mixed-precision strategy leveraging half-precision Tensor Cores (TCs) to accelerate the bottleneck back-projection (BP) kernel. edgeFBP achieves a 1.83x speedup over the widely used RTK library on Jetson Nano and a 2.56x speedup on Jetson AGX. Under a strict 25-Watt power budget, edgeFBP on Jetson Nano achieves up to 5-48x higher energy efficiency than an Nvidia DGX A100, enabling datacenter-scale imaging on constrained edge devices.