Abstract— Cardiovascular diseases are one of the most common causes of death in elderly patients. The etiology of cardiomyopathies is difficult to discern clinically. The objective of this study was to classify cardiomyopathy patients using coupling analysis, through their cardiovascular behavior and the baroreflex response. A total of thirty-eight cardiomyopathy patients (CMP) classified as ischemic (ICM, 25 patients) and dilated (DCM, 13 patients) were analyzed. Thirty elderly control subjects (CON) were used as reference. Their electrocardiographic (ECG) and blood pressure (BP) signals were studied. To characterize the cardiovascular activity, the following temporal series were extracted: beat-to-beat intervals (from the ECG signal), and end– systolic and diastolic blood pressure amplitudes (from the BP signal). Non-linear characterization techniques like high resolution joint symbolic dynamics, segmented Poincaré plot analysis, normalized short- time partial directed coherence, and dual sequence method were used to characterize these times series. The best indices were used to build support vector machine models for classification. The optimal model for ICM versus DCM patients achieved 84.2% accuracy, 76.9% sensitivity, and 88% specificity. When CMP patients and CON subjects were compared, the best model achieved 95.5% accuracy, 97.3% sensitivity, and 93.3% specificity. These results suggest a disfunction in the baroreflex mechanism in cardiomyopathies patients. I. INTRODUCTION Heart failure is one of the most common conditions suffered by elderly patients, due to age-related changes of the cardiovascular system. The etiology of these cardiomyopathies is often difficult to discern clinically. Differentiation between ischemic and dilated cardiomyopathy etiology has implications in the therapy, treatment, and prognosis of the patient. * Research supported in part by the Secretariat of Universities and Research of the Department of Economy and Knowledge of the Government of Catalonia (Consolidated research group GRC 2017 SGR 1770), by CERCA Programme/Generalitat de Catalunya, by the Spanish Ministry of Economy and Competitiveness through project DPI2015- 68820-R (MINECO/FEDER). J. Rodriguez is with Institute for Bioengineering of Catalonia (IBEC), Automatic Control Dept. (ESAII), Universitat Politècnica de Catalunya (UPC), Spain. A. Voss and S. Schulz are with the Ernst-Abbe-Hochschule Jena, University of Applied Sciences Jena, Institute of Innovative Health Technologies, Carl-Zeiss- Promenade 2. 07745 Jena, Germany (corresponding author to provide phone: 49-3641-205-625; fax: 49-3641- 205-626; e-mail: steffen.schulz@eah-jena.de, andreas.voss@eah-jena.de). B.F. Giraldo is with Automatic Control Dept. (ESAII), the Barcelona East School of Engineering (EEBE), Universitat Politècnica de Catalunya (UPC), Institute for Bioengineering of Catalonia (IBEC) – The Barcelona Institute of Science and Technology, and CIBER de Bioengeniería, Biomateriales y Nanomedicina (CIBER-BBN), Av. Eduard Maristany, 16. 08019, Barcelona, Spain. (beatriz.giraldo@upc.edu). The cardiovascular interactions are typically complex by nature, and non-linear approaches are more suited to study their dynamics. Several methods have been explored for the quantitative analysis of the cardiovascular system [1, 2]. The study of the cardiovascular couplings can bring information about the baroreflex mechanism, quantifying their direct and indirect relationships under different pathological conditions. Some studies have applied coupling methods successfully on the analysis of similar interactions [3, 4]. In our previous work, we analyzed the cardiovascular interactions in cardiomyopathy patients using different univariate linear and non-linear methods [5-7]. In this study, we propose this analysis using coupling-based non-linear methods. Also, the application of support vector machines method for classification of the patients by their etiology. II. DATABASE The noninvasive electrocardiographic (ECG) and blood pressure (BP) signals were recorded from 38 cardiomyopathy patients (CMP) at the Santa Creu i Sant Pau Hospital in Barcelona, Spain (Fig.1.). All patients were studied according to a protocol, previously approved by the Hospital ethics committee. Twenty-five of these patients had a diagnosis of ischemic cardiomyopathy (ICM) while the rest (13) had been diagnosed with dilated cardiomyopathy (DCM) disease. Thirty elderly healthy subjects (68.3 ± 6.4 years) were used as a control group (CON). The clinical information is summarized in Table 1. The recorded data was acquired with the Portapres- system and the Porti 16-biosignal amplifier for 15 minutes, at a sample frequency of 1600 Hz, with the patient in a supine position. Figure 1. Excerpt of a) ECG signal, b) BP signal III. METHODS In the ECG and BP signals the linear trend were removed. Afterwards, the customized pre-processing tools were used to reduce artifacts and spikes, and the outliers were removed. Cardiovascular Coupling-Based Classification of Ischemic and Dilated Cardiomyopathy Patients * Javier Rodriguez, Steffen Schulz, Andreas Voss, and Beatriz F. Giraldo, Senior Member, IEEE