A New Database of Multimodal Sensor Records and Echocardiographic and Peptide Test Results of HF Patients and Controls: Possibilities and Perspectives

Jovana Petrovic1, Masa Tiosavljevic2, Mirjana Stojanovic1, Aleksandar Lazovic3, Marija Ivanovic4, Aleksandra Maluckov1, Predrag Tadić5, Ljupčo Hadžievski1, Arsen D. Ristic6, Vladan D Vukcevic7
1Vinca Institute of Nuclear Sciences, University of Belgrade, 2University of Belgrade, School of Electrical Engineering, 3Vinča Institute of Nuclear Sciences, University of Belgrade, 4Vinca Institute of Nuclear Sciences, 5University of Belgrade - School of Electrical Engineering, 6Faculty of Medicine, University of Belgrade, 7Clinical Center of Serbia

Abstract

An essential prerequisite for the successful treatment of heart failure (HF) is its early diagnosis. It has been shown that the multimodal sensing devices may be capable of timely detection of HF with reduced ejection fraction (HFrEF). Recent studies indicated that the biomarkers of HF with preserved EF (HFpEF) may also be accessible to such devices. To test the possibility of HF detection by a multimodal sensor system, we have constructed a device that synchronously records an ECG, heart sounds, arterial pulsations and seismocardiogram, and are currently testing this hypothesis in a clinical study SensSmart at the University Clinical Centre of Serbia (ethical approval no. 576/15).

The secondary objective of the study is to create a database containing multimodal recordings, echocardiographic and natriuretic peptide test results, patient HF and general health history. A patient record is composed of:

• five 30-sec recordings with 10 channels each (ECG leads I, II, V1-V6, two seismograms of the chest, stethoscope, and plethysmograms of the carotid and brachial artery),

• echocardiographic results: EF, the status of the left ventricle, valves and myocardium,

• NT-proBNP values,

• case history: the presence of arterial hypertension, coronary artery disease, arrhythmias, type 2 diabetes, thyroid disease, dyslipidemia, renal failure, autoimmune diseases, previous surgeries, chronic lung disease and pulmonary embolism,

• NYHA, Killip and HFr/m/pEF ('m' stands for mid-range) classifications of HF, predominant symptoms and years since diagnosis,

• patient age, sex, height and weight.

Currently, the database contains 47 HF patient and 51 control records. At the conference, we will report an interim status of the database and the correponding distributions of recordings over HF classes, arrhythmias, valve diseases, etc. Our aim is to discuss and obtain feedback from the audience on the structuring and possible uses of the new database prior to its publication.