144 THE POTENTIAL OF SUBJECT-ADAPTIVE CENTRAL SYSTOLIC BLOOD PRESSURE ESTIMATION USING A RECURSIVE LOWPASS FILTER AND MULTIPLE LINEAR REGRESSION

Abstract

Objectives: Central blood pressure(CBP) is strongly related to cardiovascular diseases(1–3). The aim of this study was to develop a novel algorithm that was subject-adaptive for central systolic blood pressure(CSBP) and confirm its potential significance. Design and methods: We analyzed the data from 2,048 subjects (aged 47–17, 1177 males, 871 females) with SphygmoCor(AtCor, AU) measurements from the Cardiovascular hospital at Yonsei University. A recursive lowpass filter(RLPF) was used to estimate the CSBP, and its equation is as follows: eCBP(t)=α×eCBP(t-1)+(1-α)×PBP(t), where eCBP(t) and PBP(t) are the estimated CBP and peripheral blood pressure in time(t); α is the filter-coefficient which affects the filter characteristics. We calculated the optimal α which has a minimum difference between the CSBP in the SphygmoCor(GTF-CSBP) and the eCSBP, and the prediction equation for α was developed through multiple linear regression(forward stepwise regression) with subject and PBP information. The equation is as follows: αest=1.046+0.001×(0.078×Pulse area/HR-33.038×Harmonic ratio-0.936×PP-0.145×Age +0.623×Weight+0.103×pAIx-1.605×BMI-0.423×Height+1.056×PSBP-0.012×Pulse area-0.44×HR+24.452×PP/HR). The parameters were the Harmonic ratio(1st harmonic/(2nd+3rd +4th+5th harmonic), heart-rate(HR), pulse pressure(PP), body mass index(BMI), peripheral AIx(pAIx), and peripheral systolic blood pressure(PSBP). We applied the αest to the RLPF, and the performance of the proposed method was evaluated through the correlation and Bland-Altman plot between GTF-CSBP and eCSBP. Results: The proposed method had a good correlation coefficient of 0.996(p<0.001) with GTF-CSBP and a difference between GTF-CSBP and eCSBP of -0.023±0.874 mmHg. Comparison of GTF-CSBP and eCSBP showed agreement in the Bland-Altman plots. Conclusions: This study shows that the CSBP can be estimated with the proposed subject-adaptive algorithm.References Roman MJ, Devereux RB, Kizer JR, Lee ET, Galloway JM, Ali T, et al. Central pressure more strongly relates to vascular disease and outcome than does brachial pressure: the Strong Heart Study. Hypertension. 2007;50(1):197–203. Epub 2007/05/09. Williams B, Lacy PS, Thom SM, Cruickshank K, Stanton A, Collier D, et al. Differential impact of blood pressure-lowering drugs on central aortic pressure and clinical outcomes: principal results of the Conduit Artery Function Evaluation (CAFE) study. Circulation. 2006;113(9):1213–25. Epub 2006/02/16. Williams B, Lacy PS. Central haemodynamics and clinical outcomes: going beyond brachial blood pressure? Eur Heart J. 2010;31(15):1819–22. Epub 2010/05/18.