Title: A novel technique to detect cardiac function by analyzing air-flow to fingertip-oxygen lag time on polysomnography in patients with sleep disordered breathing and heart failure
Abstract: Patients with sleep disordered breathing (SDB) are 2.4 times higher predisposed to heart failure (HF) (Wang, 2007) and the prevalence of comorbid SDB rises by 76% in patients with HF (Oldenburg, 2007). Although polysomnography (PSG) is a potent tool for testing SDB, it has no power for screening HF. If it can estimate cardiac function out of PSG datasets, it would become more attractive tool for a physician treating patients with both SDB and HF. Circulation time, which indicates how long blood takes to travel a certain amount of distance in the vasculature, is one of the most important indices for estimating severity of cardiac dysfunction. In a patient with SDB, repetitive swings in SpO2 following apneic event occur with a somewhat fixed lag time (LT). This LT from air flow to finger-tip SpO2 is supposed to reflect circulation time. The aim of this study is to develop an algorithm that can detect such a LT and verify whether the LT correlates to cardiac function. We analyzed PSG data obtained from HF patients with central sleep apnea (CSA) (n = 32) and obstructive sleep apnea (OSA) (n = 23) who underwent also echocardiography in our hospital. The COMPUMEDICS E-series (COMPUMEDICS Co Ltd) was used for sleep examination. SpO2 sensor was attached to the left finger-tip. In signal processing, we full-rectified the airflow signals and applied low-pass filter with 0.5 Hz cutoff frequency to the airflow and SpO2 signals. These data were analyzed using cross-correlation algorithm and LT was determined. We examined correlation between LT and left ventricular ejection fraction (LVEF). Data processing and statistical analysis were performed with MATLAB 2007a (MathWorks Inc) and Excel 2010 (Microsoft Corp), respectively. Patients background were age; 64.2 ± 14.3 year-old, apnea hypopneaindex; 46.2 ± 30.1/h, LVEF; 55.9 ± 19.0%. With the cross-correlation algorithm, LT was robustly detected at every apnea/hypopnea events. As compared with OSA to CSA, significant difference in LVEF and LT were found (OSA vs. CSA, LVEF; 66.8 ± 7.6% vs. 48.1 ± 20.7%, p = 0.0001, LT; 28.6 ± 3.8 s vs. 36.4 ± 7.4 s, p < 0.0001). Overall scatter plots shows significant linear correlation between LT and LVEF (LVEF = 3.2 * (50.3-LT), R2 = 0.64, p < 0.0001). Our novel analysis algorithm using data of usual PSG can be a simple and useful tool for screening HF and estimating cardiac function. We thank Tanaka Y, Tanaka K for data acquisition.
Publication Year: 2013
Publication Date: 2013-12-01
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot