Programa del congreso

This work presents an in-vehicle monitoring system based on a PCR (Pulsed Coherent Radar) at the 60 GHz ISM unlicensed band. Since the radar can measure distances and displacements with sub-millimeter resolution, a system for seat occupancy detection and vehicle occupants’ gesture recognition and breathing monitoring is proposed. The measured results of breathing agree with those obtained by another system whose measurements are based on the airflow, which allows for its validation.

According to the OMS, the highest rate of death among non-communicable diseases is caused by cardiovascular disorders (CVDs), killing approximately 18 million people yearly. Early detection of arrhythmias or CVDs is critical. For this reason, new portable acquisition devices have made it possible the recording of electrocardiograms (ECGs) for extended periods of time. This is the most common method for identifying arrhythmias with a random nature. However, the acquisition environment may cause distortions or even destroy the ECG recordings, making difficult to diagnose CVDs accurately. As a result, the evaluation of the ECG signal quality is required. The suggested method is based on feature and meta-feature extraction capabilities from 5-second ECG segments, allowing machine learning classifiers to identify between ECG segments of high and low-quality. Results showed that the utilization of meta-features increases the accuracy of the methodology, compared with the utilization of the original hand-crafted feature set.

This article presents a broadband rectangular-shaped patch antenna with coplanar feed. The antenna is composed of a ground plane and a step-shaped radiating patch. The final size of the proposed antenna is 30 x 30 mm², and it is manufactured onto a low-cost FR4 substrate. The measured results of the antenna indicate that its operating range is between 2.5 and 24.9 GHz, the gain is between 2.1 and 5.4 dBi, and it presents omnidirectional radiation patterns. The results of the antenna in the time domain are also evaluated, with a very high similarity between transmitted and received broadband pulses and high integrity of the transmitted signals. After the evaluation of the results obtained with the proposed antenna, it is shown that it is a good candidate to be used in a medical imaging system.

Cardiovascular data provides valuable information into daily health, and the increasing adoption of wearable devices, such as photoplethysmography (PPG), promises extensive diagnosis and analysis capabilities. However, validation through established clinical standards like electrocardiogram (ECG), either simultaneously or closely aligned with PPG, remains essential for validating cardiac disease diagnosis. This paper introduces a novel approach to synchronize ECG and PPG recordings based on the cardiac variability dynamics and non-linear techniques. Through a ``proof of concept" analysis on a controlled dataset of 13 patients, the efficacy of mutual information (MI) versus linear correlation (LC) is compared. The findings underscore the superior performance of MI over the traditional LC method, thereby highlighting immediate opportunities for further research on signal synchronization, particularly in scenarios involving the physical collection of cardiovascular data from different devices.