Electrocardiogram (ECG) signals are vital for diagnosing cardiovascular diseases but are often compromised by artifacts such as high-frequency noise and baseline wander (BW). This review synthesizes recent advancements in ECG enhancement techniques aimed at improving signal quality for clinical evaluation. Key methods discussed include empirical mode decomposition (EMD) and its variants, which effectively mitigate both high-frequency noise and BW with minimal distortion. Additionally, deep learning approaches have emerged as promising solutions for BW filtering, demonstrating superior performance compared to traditional methods. The review also examines techniques like eigenvalue decomposition of Hankel matrices for simultaneous noise removal, as well as integrated methods such as IEMD-ATD that enhance signal-to-noise ratio while preserving crucial features of the ECG signal. Furthermore, we explore the application of ensemble EMD and non-local wavelet transforms, both of which provide significant improvements in denoising efficacy. The proposed methodologies not only enhance the diagnostic potential of ECG recordings but also contribute to the broader context of wearable health monitoring systems.