Ultrasound Beamforming: Exploring Cloud-Native and Edge Computing Solution

Abstract

Ultrasound imaging is pivotal in modern medicine for its non-invasive visualization of anatomical structures. How- ever, traditional systems face challenges in scalability, flexibility, and cost-effectiveness due to centralized processing units. This paper introduces a novel cloud-native architecture tailored for ultrasound imaging, focusing on the computationally intensive beamforming process. The architecture integrates on-premise edge-computing acquisition probes with cloud-based processing, leveraging edge computing to enhance efficiency and responsive- ness. Initial data preprocessing at the edge optimizes bandwidth use before secure transmission to the cloud via dedicated connections. In the cloud, real-time data ingestion through AWS Kinesis supports scalable processing using containerization technologies. High-performance computing languages ensure efficient beam- forming, improving diagnostic accuracy and operational efficacy. The architecture's scalability, cost-effectiveness, and integration potential for AI-driven diagnostics are discussed, highlighting its transformative impact on ultrasound imaging practices and healthcare delivery.