Investigation of Cardiological Data By Nonlinear Canonical Correlation Analysis

Year 2002, Volume: 1 Issue: 3, 55 - 63, 16.12.2002

Mevlüt Türe , Necdet Süt , Gültaç Özbay Turhan Kürüm

Abstract

Nonlinear Canonical Correlation Analysis (NLCCA) is a method to reveal relations between two or more sets contains nominal, ordinal, numerical or mixed variables.

In this study 526 patients who were coronary angiographies were studied at Department of Cardiology, Faculty of Medicine at Trakya University from 1998 to 2000. Coronary artery diameters were taken up in the first set (circumflex: CX, Left Anterior Descending: LAD, Right Coronary Artery: RCA and Left Main Coronary Artery: LMCA), lesion in the coronary artery diameters were taken up in the second set and sex, diabetes mellitus (DM), smoking and body mass index (BMI) variables were taken up in the third set. The relations between these variables were studied with NLCCA.

Result of analysis with two-dimensional solution, the first dimension explained 54.7% and the second dimension explained 44.3% of relations between the sets of variables. The significant correlations were found between lesions in the LAD and CX with diameters of DM, RCA, and LMCA, and also between smoking, sex, lesion in the RCA and BMI, CX, LAD diameters.

Keywords

Nonlinear Canonical Correlation Analysis, Optimal Scaling, Coronary Artery

Kardiyolojik Verilerin Doğrusal Olmayan Kanonik Korelasyon Analizi ile İncelenmesi

Abstract

Doğrusal olmayan kanonik korelasyon analizi (DOKKA) isimsel, sıralı, nümerik ya da bunların karışımı değişkenlerden oluşan iki ya da daha fazla değişken seti arasındaki ilişkileri ortaya koymada yararlanılan bir yöntemdir.

Bu çalışmada Trakya Üniversitesi Tıp Fakültesi kardiyoloji kliniğinde 2000 yılına kadar koroner anjiyografisi yapılan 526 hastanın koroner arter çapları (1. Set-circumflex:CX, sol ön inen damar: LAD, sağ koroner arter: RCA ve sol ana koroner arter: LMCA (¯X ±1*s) alınarak k:küçük, o:orta, b:büyük), koroner arter çaplarında lezyon olup olmadığı (2.set), cinsiyet, diabetes mellitus (DM), sigara kullanımı ve vücut kitle indeksi (VKİ) (3.set) değişkenleri ele alınarak bu üç setteki değişkenler arasındaki ilişkiler DOKKA ile incelendi.

Analiz sonucunda iki boyutlu çözümlemede birinci boyutun değişkenler arasındaki ilişkinin %54.7’sini, ikinci boyutun ise %44.3’ünü açıkladığı ve DM, RCA ve LMCA çaplarının LAD ve CX’teki lezyonlarla; sigara, cinsiyet ve RCA’daki lezyonun VKİ, CX ve LAD çapları ile korelasyonlu olduğu bulundu.

Keywords

Doğrusal olmayan kanonik korelasyon analizi, Optimal Ölçekleme, Koroner Arter

References

• CATRIEN, C.J.H. & VAN DER BURG, E. (1993), Longitudinal K-Sets Analysis Using Lagged Variables, QUESTIIO, 17, 327-338.
• GIFI, A. (1990), Nonlinear Multivariate Analysis, Chichester, John Wiley & Sons.
• HENK, A. KIERS L. (1989), Three-Way Methods for the Analysis of Qualitative and Quantitative Two Way Data, Leiden, DSWO PRESS.
• DE LEEUW, J. (1984), Canonical Analysis of Categorical Data, Leiden, DSWO PRESS.
• NWASOKWA, O.N., WEISS, M., GLADSTONE, C., BODENHEIMER, MM. (1996), Effect of coronary artery size on the prevalence of atherosklerosis. Am J. Cardiol, 78, 741-746.
• ÖZDAMAR, K. (1999), Paket Programlar ile İstatistiksel Veri Analizi-2, Eskişehir, Kaan Kitapevi.
• SPSS INC,. (1996), SPSS Categories 10.0 Chicago. SPSS Inc.
• TATLIDİL, H. (1996), Uygulamalı Çok Değişkenli İstatistiksel Analiz, Ankara, Cem Web Ofset.
• VAN DER BURG, E. (1988), Nonlinear Canonical Correlation and Some Related Techniques, Leiden, DSWO PRESS.
• VAN DER BURG, E., DE LEEUW, J. (1987), Nonlinear Canonical Correlation Analysis with k Sets of Variables, Eric Document, Research Report 87-8.
• VANDERBURG, E., DE LEEUW, J., DIJKSTERHUIS, G. (1994), OVERALS- Nonlinear Canonical Correlation with K Sets of Variables, Computational Statistics & Data Analysis, 18(1), 141-163.
• VANDERBURG, E., DIJKSTERHUIS, G., (1996), Generalised Canonical Analysis of Individual Sensory Profiles and Instrumental Data, Multivariate Analysis of Data in Sensory Science, Elseiver Science B.V, 221-258.
• VAN DER BURG, E., DE LEEUW, J. (1983), - Nonlinear Canonical Correlation, British Journal of Mathematical and Statistical Psychology, 36, 54-80.
• VAN DER BURG, E., DIJKSTERHUIS, G., (1989), Nonlinear Canonical Analysis of Multiway Data: Multiway Data Analysis, 147-158.
• VAN DER BURG, E., DIJKSTERHUIS, G., (1993), Nonlinear Generalised Canonical Analysis: Introduction and Application from Sensory Research, Advances in Longitudinal and Multivariate Analysis in the Behavioral Sciences, ITS, 193-203.