RESUMEN
The timely detection of abnormal electrocardiogram (ECG) signals is vital for preventing heart disease. However, traditional automated cardiology diagnostic methods have the limitation of being unable to simultaneously identify multiple diseases in a segment of ECG signals, and do not consider the potential correlations between the 12-lead ECG signals. To address these issues, this paper presents a novel network architecture, denoted as Branched Convolution and Channel Fusion Network (BCCF-Net), designed for the multi-label diagnosis of ECG cardiology to achieve simultaneous identification of multiple diseases. Among them, the BCCF-Net incorporates the Channel-wise Recurrent Fusion (CRF) network, which is designed to enhance the ability to explore potential correlation information between 12 leads. Furthermore, the utilization of the squeeze and excitation (SE) attention mechanism maximizes the potential of the convolutional neural network (CNN). In order to efficiently capture complex patterns in space and time across various scales, the multi branch convolution (MBC) module has been developed. Through extensive experiments on two public datasets with seven subtasks, the efficacy and robustness of the proposed ECG multi-label classification framework have been comprehensively evaluated. The results demonstrate the superior performance of the BCCF-Net compared to other state-of-the-art algorithms. The developed framework holds practical application in clinical settings, allowing for the refined diagnosis of cardiac arrhythmias through ECG signal analysis.
Asunto(s)
Algoritmos , Cardiología , Humanos , Redes Neurales de la Computación , Arritmias Cardíacas/diagnóstico , Electrocardiografía/métodosRESUMEN
Data collection problems have received much attention in recent years. Many data collection algorithms that constructed a path and adopted one or more mobile sinks to collect data along the paths have been proposed in wireless sensor networks (WSNs). However, the efficiency of the established paths still can be improved. This paper proposes a cooperative data collection algorithm (CDCA), which aims to prolong the network lifetime of the given WSNs. The CDCA initially partitions the n sensor nodes into k groups and assigns each mobile sink acting as the local mobile sink to collect data generated by the sensors of each group. Then the CDCA selects an appropriate set of data collection points in each group and establishes a separate path passing through all the data collection points in each group. Finally, a global path is constructed and the rendezvous time points and the speed of each mobile sink are arranged for collecting data from k local mobile sinks to the global mobile sink. Performance evaluations reveal that the proposed CDCA outperforms the related works in terms of rendezvous time, network lifetime, fairness index as well as efficiency index.
RESUMEN
Gaussian modelling method has been reported as a useful method to analyze arterial pulse waveform changes. This study aimed to provide scientific evidence on Gaussian modelling characteristics changes derived from the finger photoplethysmographic (PPG) pulses during exercise and recovery. 65 healthy subjects (18 female and 47 male) were recruited. Finger PPG pulses were digitally recorded with 5 different exercise loads (0, 50, 75, 100, 125W) as well as during each of 4minute (min) recovery period. The PPG pulses were normalized in both width and amplitude for each recording, which were decomposed into three independent Gaussian waves with nine parameters determined, including the peak amplitude (H1, H2, H3), peak time position (N1, N2, N3) and half-width (W1, W2, W3) from each Gaussian wave, and four extended parameters determined, including the peak time interval (T1,2, T1,3) and amplitude ratio (R1,2, R1,3) between 1st Gaussian wave and 2nd, 3rd Gaussian waves. These derived parameters were finally compared between different exercise loads and recovery phases. With gradually increased exercise loads, the peak amplitude H2, peak time position N1, N2, N3, and half-width W1, W2 increased, peak amplitude H3 decreased significantly (all P<0.05). The peak time interval T1,2 and T1,3 increased significantly from 10.6±1.2 and 36.0±4.4 at rest to 14.4±2.3 and 45.1±6.5 at 100W exercise load, respectively (both P<0.05). The amplitude ratio R1,2 also increased from 1.07±0.2 at rest to 1.22±0.2 at 100W, and the amplitude ratio R1,3 decreased from 1.10±0.3 at rest to 0.42±0.2 at 125W (all P<0.05). An opposite changing trend of these parameters was observed during recovery phases. In conclusion, this study has quantitatively demonstrated significant changes of Gaussian modelling characteristics derived from finger PPG pulse with exercise and during recovery, providing scientific evidence for the physiological mechanism that exercise increases cardiac ejection and vasodilation, and reduces the total peripheral vascular resistance.
Asunto(s)
Ejercicio Físico/fisiología , Dedos/irrigación sanguínea , Hemodinámica , Modelos Cardiovasculares , Fotopletismografía/métodos , Análisis de la Onda del Pulso/métodos , Procesamiento de Señales Asistido por Computador , Adulto , Presión Arterial , Gasto Cardíaco , Femenino , Humanos , Masculino , Valor Predictivo de las Pruebas , Recuperación de la Función , Factores de Tiempo , Resistencia Vascular , Vasodilatación , Adulto JovenRESUMEN
It is physiologically important to understand the arterial pulse waveform characteristics change during exercise and recovery. However, there is a lack of a comprehensive investigation. This study aimed to provide scientific evidence on the arterial pulse characteristics change during exercise and recovery. Sixty-five healthy subjects were studied. The exercise loads were gradually increased from 0 to 125 W for female subjects and to 150 W for male subjects. Radial pulses were digitally recorded during exercise and 4-min recovery. Four parameters were extracted from the raw arterial pulse waveform, including the pulse amplitude, width, pulse peak and dicrotic notch time. Five parameters were extracted from the normalized radial pulse waveform, including the pulse peak and dicrotic notch position, pulse Area, Area1 and Area2 separated by notch point. With increasing loads during exercise, the raw pulse amplitude increased significantly with decreased pulse period, reduced peak and notch time. From the normalized pulses, the pulse Area, pulse Area1 and Area2 decreased, respectively, from 38 ± 4, 61 ± 5 and 23 ± 5 at rest to 34 ± 4, 52 ± 6 and 13 ± 5 at 150-W exercise load. During recovery, an opposite trend was observed. This study quantitatively demonstrated significant changes of radial pulse characteristics during different exercise loads and recovery phases.