This chapter presents a comparative study of artificial intelligence (AI)-based binary classifiers designed to assess the integrity of electrocardiogram (ECG) data, with a particular focus on implementation at the network edge. Ensuring data integrity at the edge allows for the early identification and exclusion of corrupted or low-quality signals, thereby reducing the volume of data that must be stored or transmitted. This, in turn, lowers the power and storage demands on Internet-of-Things (IoT) devices without compromising system performance. The chapter investigates various machine learning classifiers that evaluate ECG signal quality using low-complexity Signal Quality Indices (SQIs) derived from kurtosis and skewness measures. Experimental results reveal that a bagged ensemble of three neural networks offers the highest accuracy in detecting poor-quality ECG signals. Additionally, the chapter analyzes the computational complexity and energy consumption of individual classifier models as well as classifier fusion approaches, providing practical insights for resource-constrained environments. Together, these findings support the development of efficient, AI-driven edge solutions for real-time biomedical signal validation.