Comparative Analysis of Edge-Enabled AI–IoT Healthcare Systems: Toward the Most Effective Predictive Model

Review

Abstract

The combination of the Artificial Intelligence (AI), the Internet of Things (IoT), and the Internet of Medical Things (IoMT) is driving current developments in the field of intelligent healthcare systems, thus, enabling the continuous, real-time, and remote monitoring of patients. The traditional cloud-based healthcare infrastructure, however, is riddled with latency, bandwidth, scaling, and increased risks to the privacy of data, all of which decrease their efficiency in time-intensive clinical settings. To address these drawbacks, this study performs a comparative analysis of edge-based AI-IoT healthcare systems, in which the most useful predictive model is determined to be deployed on clinical scenarios in real time. Various machine-learning and deep-learning frameworks have been tested on hybrid edge-cloud systems using ECG data gathered by IoT and the performance indices included prediction accuracy, latency, computation efficiency, and edge-feasibility. Table 1 enumerates the results of applying the convolutional neural network (CNN)-based models, showing that they have higher performance, achieving around 99 percent success on arrhythmia detection, with low latency that makes them an option of implementing them in the form of edge devices. Comparatively, a three-layer-based monitoring setup based on the concept of combining IoT devices with a hybrid CNN-UUGRU model achieves 97.7 percent accuracy on public datasets and can be used to partition edge-cloud tasks and mobile-based notifications, but at slightly reduced predictive accuracy. Although benchmark results are strong, the enduring weaknesses are linked to reliance on clean and non-clinical datasets and lack of a thorough test of robustness, privacy, and extended deployment. Comprehensively, it can be indicated that CNN-based models provide the best balance in terms of accuracy, real-time performance, and edge feasibility; thus, its significant potential in scalable, patient-centered, and reliable edge-enabled AI-IoT healthcare systems, especially regarding real-time arrhythmia screening.