NXP Semiconductors and GE HealthCare announced at this week’s CES in Las Vegas a collaboration to pioneer advancements in edge AI innovation that help clinicians improve patient care with simple, actionable insights.

Leveraging NXP’s history in secure edge processing and GE HealthCare’s experience in medical technology, the collaboration is beginning with anaesthesia and neonatal concepts showcased at CES, highlighting how the intelligence, low latency, resilience and security offered by edge AI can transform certain workflows.

In acute care environments such as the operating room and neonatal intensive care unit (NICU), care teams want actionable intelligence delivered with low latency. On-device edge AI processes data to deliver immediate, consistent performance and help enable critical insights to be available without delay. Embedding edge AI into point-of-care devices may provide clinicians fast, reliable and secure access to critical information.

GE HealthCare and NXP have formulated two concepts combining NXP’s edge AI expertise and portfolio with GE HealthCare’s experience in medical technology innovations. The first brings edge AI capabilities to anaesthesia delivery in the operating room, intending to provide anaesthesiologists with a reliable, hands-free way to interact with anaesthesia equipment using real-time voice commands. This project is designed to help anaesthesiologists attend to the patient in a dynamic and crowded operating room, while reducing excessive cognitive load, alarm fatigue and risk of human error.

The second concept is designed to support neonatal care through intelligent, live monitoring. This technology aims to detect whether an infant is crying or at rest, identify unwanted objects in the bed, or recognise if the baby has rolled onto their stomach, designed to help care teams and enable them to keep babies comfortable. This concept uses agentic AI on the edge to log the event and alert clinicians, if appropriate. All image processing occurs locally using models leveraging the NXP eIQ AI toolkit (www.nxp.com/design/design-center/software/eiq-ai-development-environment:EIQ). No images leave the device, supporting strict security and privacy standards.

The development of both concepts is underpinned by GE HealthCare’s responsible AI principles (www.gehealthcare.com/insights/article/responsible-ai-in-healthcare-from-theory-to-practice), which include a focus on safety, security, privacy, validity, transparency, explainability and fairness. The concepts are based on NXP’s applications processors with integrated neural processing units (NPUs), as well as a dedicated, standalone NPU, with software enablement from NXP’s eIQ AI toolkit.

“At GE HealthCare, we build AI that keeps clinicians at the centre, assisting clinical judgment and freeing up time for patient care,” said Jeff Caron, chief technology officer at GE HealthCare (www.gehealthcare.com). “Collaborating with NXP helps us explore secure on-device AI as a complement to our cloud, with concepts designed to support care teams in acute settings.”

Charles Dachs, executive vice president at NXP (www.nxp.com), added: “This collaboration brings together GE HealthCare’s clinical trust and decades of medical technology innovation with NXP’s eIQ AI enablement and deep experience in secure, high-performance edge computing to provide safe, secure and practical edge AI to clinicians and patients. Together, we aim to enable more personalised care, from continuous monitoring in the NICU and hands-free interaction with anaesthesia equipment to exploratory research concepts such as AI-driven risk prediction, automated triage and personalised treatment recommendations.”