Edge AI for Internet of Medical Things: Optimized Algorithms and Architectures for Medical Diagnostics

Traditional healthcare AI heavily relies on centralized cloud computing, which introduces critical latency, high bandwidth costs, and substantial data privacy risks when handling sensitive patient information. Edge AI offers a transformative alternative by shifting intelligence directly to local medical devices, wearables, and clinical hardware. This paradigm shift enables split-second diagnostic insights at the point of care, ensures continuous monitoring in remote or low-connectivity areas, and inherently enhances data privacy by keeping patient information localized.As the Internet of Medical Things (IoMT) rapidly expands, the healthcare sector faces an urgent need to bridge the gap between complex, resource-heavy AI architectures and the strict resource constraints of medical hardware. Exploring efficient, decentralized AI is essential to unlocking the next generation of smart, resilient, and ubiquitous healthcare diagnostics.The primary goal of this Research Topic is to aggregate pioneering, high-quality research that addresses the core challenges of deploying artificial intelligence directly at the healthcare edge. While the potential of Edge AI in diagnostics is immense, bridging the gap between sophisticated deep learning models and resource-constrained medical hardware requires significant innovation in algorithmic efficiency, hardware acceleration, and decentralized frameworks.This collection aims to foster interdisciplinary collaboration among computer scientists, biomedical engineers, and clinicians to drive the evolution of smart medical devices. We seek to highlight novel methodologies that optimize model size and computational speed without sacrificing clinical accuracy. Furthermore, a key objective is to advance privacy-preserving architectures, ensuring that localized intelligence complies with stringent global healthcare data regulations. Ultimately, this topic intends to establish a comprehensive roadmap for transitioning theoretical Edge AI models into validated, safe, and scalable clinical tools that enhance patient outcomes worldwide.This Research Topic welcomes original research, and reviews, focused on the theory, design, implementation, and validation of Edge AI systems for healthcare. We encourage submissions exploring, but not limited to, the following themes:- Lightweight Architectures: Model compression, pruning, and quantization optimized for resource-constrained medical devices.- Privacy-Preserving AI: Federated learning frameworks and decentralized architectures for secure clinical data processing.- On-Device Diagnostics: Smart wearables, point-of-care devices, and remote patient monitoring applications.- Hardware-Software Co-Design: Energy-efficient computing and neuromorphic hardware tailored for the IoMT.- Real-Time Analytics: Edge processing of biosignals, high-resolution medical imagery, and continuous telemetry. Topic editors: Mohamed Shaban, Mohamed El-Sharkh, Edmund Spencer, Ahmed Abdelgawad Source: https://www.frontiersin.org/research-topics/82635/edge-ai-for-internet-of-medical-things-optimized-algorithms-and-architectures-for-medical-diagnostics