Electrical and Electronic Engineering - Research Publications
Permanent URI for this collection
Search Results
1 - 10 of 10
-
Item
-
Item
-
Item
-
ItemInteraction between sleep EEG and ECG signals during and after obstructive sleep apnea events with or without arousals(IEEE - Institute of Electrical and Electronic Engineers, 2008)
-
ItemIdentification of onset, maximum and termination of obstructive sleep apnoea events in single lead ECG recordings(IEEE, 2008)Measuring the Apnoea Hypopnoea Index (AHI) is important for determining the severity of any apnoea patient. This study presents a method of screening each apnoea event separately based on the single lead Electrocardiogram (EGG) signal. The whole ECG of a subject was divided into Normal, Onset, OSA-maximum and Termination epochs with length of 5 seconds. PSD analysis was used for determining the features directly from the ECG. ROC area was calculated to determine the discrimination capability of each feature (or power in each frequency bin) found by PSD analysis. The maximum ROC area found between Normal vs. OSA-maximum was 0.81 in the frequency range of 52-72 Hz. The ROC area and significant frequency band for Normal vs. Onset and Normal vs. Termination were 0.78, 0.78 and 57-65 Hz, 52-66 Hz respectively.
-
ItemAnalysis of coherence between sleep EEG and ECG signals during and after obstructive sleep apnea events(IEEE, 2008)This study presents the first successful preliminary attempt to directly investigate the interactions of power spectra of sleep EEG and ECG signals of patients with obstructive sleep apnea syndrome (OSAS) by coherence analysis. ECG and EEG signals were collected from 8 OSAS patients and 3 healthy subjects. Coherence between two signals over different frequency bands(0-128 Hz) were calculated for normal breathing events, obstructive sleep apnea (OSA) events and events following OSA terminations (with/without arousals) in non-REM as well as REM sleep. Overall coherence of ECG and EEG in REM sleep is higher than that in non-REM sleep. A significant (p=0.0164) difference of coherence in the range of 10-5 Hz was found among normal, OSA and termination events in REM sleep. The results could be useful in detecting OSA events or OSA related arousals to characterize sleep fragmentation from ECG and EEG signals.
-
Item
-
Item
-
ItemInvestigating scale invariant dynamics in minimum toe clearance variability of the young and elderly during treadmill walking(IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2008-08)Current research applying variability measures of gait parameters has demonstrated promise for helping to solve one of the "holy grails" of geriatric research by defining markers that can be used to prospectively identify persons at risk of falling . The minimum toe clearance (MTC) event occurs during the leg swing phase of the gait cycle and is a task highly sensitive to the spatial and balance control properties of the locomotor system. The aim of this study is to build upon the current state of research by investigating the magnitude and dynamic structure from the MTC time series fluctuations due to aging and locomotor disorder. Thirty healthy young (HY), 27 healthy elderly (HE), and 10 falls risk (FR) elderly individuals (who presented a prior history of trip-related falls) participated in treadmill walking for at least 10 min at their preferred speed. Continuous MTC data were collected and the first 512 data points were analyzed. The following variability indices were quantified: 1) MTC mean and standard deviation (SD), 2) PoincarE plot indices of MTC variability (SD1, SD2, SD1/SD2), 3) a wavelet based multiscale exponent beta to describe the dynamic structure of MTC fluctuations, and 4) detrended fluctuation analysis exponent alpha to investigate the presence of long-range correlations in MTC time series data. Results showed that stride-to-stride MTC time series has a nonlinear structure in all three groups when compared against randomly shuffled surrogate MTC data. Test on aging effects showed the MTC central tendency was significantly lower (p < 0.01) and the magnitude of the MTC variability significantly higher (p < 0.01). This trend changed when comparing FR subjects against age-matched HE as both the central tendency (p < 0.01) and magnitude of the variability (p < 0.01) increased significantly in FR. Although the magnitude of MTC variability increased with age, the nonlinear indices represented by alpha, beta, and SD1/SD2 demonstrated that the nonlinear structure of MTC does not change significantly due to aging (p > 0.05). There were, however, significant differences between HY and FR for beta (between scale 1 and 2; p < 0.01) and alpha (p < 0.05). Out of all the variability measures applied, beta(Wv2-4), SD1/SD2, SD2 of critical MTC parameter were found to be potential markers to be able to reliably identify FR from HE subjects. Further research is required to understand the mechanisms underlying the cause of MTC variability.
-
ItemSupport Vector Machines for Automated Recognition of Obstructive Sleep Apnea Syndrome From ECG Recordings(IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2009-01)Obstructive sleep apnea syndrome (OSAS) is associated with cardiovascular morbidity as well as excessive daytime sleepiness and poor quality of life. In this study, we apply a machine learning technique [support vector machines (SVMs)] for automated recognition of OSAS types from their nocturnal ECG recordings. A total of 125 sets of nocturnal ECG recordings acquired from normal subjects (OSAS - ) and subjects with OSAS (OSAS +), each of approximately 8 h in duration, were analyzed. Features extracted from successive wavelet coefficient levels after wavelet decomposition of signals due to heart rate variability (HRV) from RR intervals and ECG-derived respiration (EDR) from R waves of QRS amplitudes were used as inputs to the SVMs to recognize OSAS +/- subjects. Using leave-one-out technique, the maximum accuracy of classification for 83 training sets was found to be 100% for SVMs using a subset of selected combination of HRV and EDR features. Independent test results on 42 subjects showed that it correctly recognized 24 out of 26 OSAS + subjects and 15 out of 16 OSAS - subjects (accuracy = 92.85%; Cohen's kappa value of 0.85). For estimating the relative severity of OSAS, the posterior probabilities of SVM outputs were calculated and compared with respective apnea/hypopnea index. These results suggest superior performance of SVMs in OSAS recognition supported by wavelet-based features of ECG. The results demonstrate considerable potential in applying SVMs in an ECG-based screening device that can aid a sleep specialist in the initial assessment of patients with suspected OSAS.