Artificial Intelligence of Medical Things (AIoMT) requires storing, preprocessing, monitoring, and analytics of large-scale sensor data fusion in the cloud. However, migrating to the cloud possesses intrinsic issues of cost, performance constraints, and sustainable computing. This research explores the potential of AIoMT in crafting intelligent models for daily activity patterns and predicting unusual occurrences. It delves into power-efficient and sustainable computing tailored for the IoT sensors, methods, and systems geared toward crafting digitally enhanced smart homes for the elderly. Fusion data is collected from heterogenous sensors to track daily patterns and processed for anomaly detection and alert generation. The AIoMT model has employed the time and energy minimization scheduler (TEMS) algorithm, which considers energy consumption, processing duration, data transmission expenses, and standby device power consumption. This enables local computing in the IoMT systems, mobile edge servers, and cloud controllers, promoting sustainability in healthcare. To optimize execution time and cost-effectiveness, task scheduling options include local Internet of Things devices, cloud infrastructure, and multiaccess edge computing (MEC). This approach could benefit digitally enhanced communities significantly, promoting low-carbon, power-efficient, sustainable computing (LCPESC). The LCPESC AIoMT approach demonstrates precision close to a 95% confidence level. Further, the proposed model is extended beyond individual households to encompass digitally augmented communities.