Daniel Proano

Abstract
We introduce a multimodal Signal Quality Indicator (SQI) for assessing fidelity of synchronous electrocardiogram (ECG) and phonocardiogram (PCG) signals recorded in ambulatory, non-standardized settings. The method uses a bidirectional fiducial-matching algorithm to test the temporal alignment of QRS complexes and T waves (ECG) with S1 and S2 sounds (PCG) respectively. Validation employed 564 synchronous ECG–PCG pairs collected with the FDA-cleared Rijuven Cardiosleeve at the aortic, pulmonary, tricuspid, and mitral valves sites. Expert annotations served as ground truth. In a three-class task, the SQI reached an area under the ROC curve greater than 79%, showing strong discriminative power. This physiology-based metric supports batch-online monitoring and reliable quality control of opportunistic cardiac recordings.

Abstract
Accurate segmentation of heart sound signals (phonocardiograms, PCGs) is a critical step for the early diagnosis of cardiovascular diseases (CVDs). Although deep learning models, particularly convolutional neural networks (CNNs) like the UNet, have achieved strong performance in PCG segmentation, the impact of signal preprocessing remains underexplored. In this study, we evaluate how different preprocessing strategies, namely wavelet-based denoising, Butterworth filtering, and their combination, affect the segmentation performance of a 1 D UNet model. Using the PhysioNet 2016 database, we evaluated segmentation quality based on sample accuracy, positive predictive value, and sensitivity. The results show that minimal preprocessing, specifically Butterworth bandpass filtering alone, yields the best segmentation performance, outperforming more aggressive preprocessing pipelines. These findings highlight that preserving the baseline structure of PCG signals is crucial for optimal learning and that lightweight preprocessing remains an essential consideration, especially when applying modern deep learning architectures. © 2025 IEEE.

Abstract
In recent times, Artificial Intelligence (AI) has become ubiquitous in technological fields, mainly due to its ability to perform computations in distributed systems or the cloud. Nevertheless, for some applications -as the case of EMG signal processing- it may be highly advisable or even mandatory an on-the-edge processing, i.e., an embedded processing methodology. On the other hand, sEMG signals have been traditionally processed using LTI techniques for simplicity in computing. However, making this strong assumption leads to information loss and spurious results. Considering the current advances in silicon technology and increasing computer power, it is possible to process these biosignals with Al-based techniques correctly. This paper presents an embedded-processing-based adaptive filtering system (here termed edge AI) being an outstanding alternative in contrast to a sensor-computer- actuator system and a classical digital signal processor (DSP) device. Specifically, a PYNQ-Z1 embedded system is used. For experimental purposes, three methodologies on similar processing scenarios are compared. The results show that the edge Al methodology is superior to benchmark approaches by reducing the processing time compared to classical DSPs and general standards while maintaining the signal integrity and processing it, considering that the EMG system is not LTI. Likewise, due to the nature of the proposed architecture, handling information exhibits no leakages. Findings suggest that edge computing is suitable for EMG signal processing when an on-device analysis is required.

Abstract

Abstract