Katalog Plus
Bibliothek der Frankfurt UAS
Bald neuer Katalog: sichern Sie sich schon vorab Ihre persönlichen Merklisten im Nutzerkonto: Anleitung.
Dieses Ergebnis aus BASE kann Gästen nicht angezeigt werden.  Login für vollen Zugriff.

Multimodal Sensing For AI-Assisted Diagnoses of Heart Failure

Title: Multimodal Sensing For AI-Assisted Diagnoses of Heart Failure
Authors: Stojanović, Mirjana; Tiosavljević, Maša; Lazović, Aleksandar; Atanasoski, Vladimir; Tadić, Predrag; Ivanović, Marija; Hadžievski, Ljupčo; Maluckov, Aleksandra; Ristić, Arsen; Vukčević, Vladan; Petrović, Jovana
Source: EECSS25 : 11th World Congress on Electrical Engineering and Computer Systems and Sciences : Proceedings
Publication Year: 2025
Collection: VinaR Repository (Vinča Institute of Nuclear Sciences / University of Belgrade)
Description: Motivated by the challenge of finding a screening-compatible method for early detection of heart failure (HF), we have revised polycardiography using the modern sensing technology. Polycardiography is a multimodal measuring technique based on the synchronous noninvasive recording of the electrical and mechanical parameters of the cardiovascular system [1]. Our prototype polycardiograph simultaneously acquires an electrocardiogram (ECG), phonocardiogram (PCG), seismocardiograms (SCG) and photoplethysmograms (PPG) [2]. It provides a direct access to electro-mechanical biomarkers of HF with reduced ejection fraction (HFrEF), notably the systolic-time intervals such as left ventricular ejection time (LVET) and the pre-ejection period (PEP). Moreover, there are strong indications that it can provide a sufficient set of biomarkers for conclusive diagnostics of HF with preserved ejection fraction (HFpEF). We are currently performing a clinical study SensSmart at the University Clinical Centre of Serbia, in which we are testing the non-inferiority of the HF diagnosis by polycardiography with respect to the echocardiography. Preliminary results of AI-based classification preformed on the median heartbeats from a set of 100 HF patients and controls, indicate a possibility to diagnose both HFrEF and HFpEF. At the conference, we will present the interim AI-based results and the study of the significance of particular noninvasively accessible features to HF classification. The clinical study SensSmart was preceded by a validation study SensSmartTech performed on healthy volunteers. Besides the technical validation of the prototype, its goal was detection of the HF biomarker (LVET, PEP, EF) dependence on heart rate (HR). Thus, the volunteers were measured before and after running on a treadmill, producing consistent heart relaxation signals over a wide range of HRs. The SensSmartTech database provides the valuable information on the complex nonlinear heart relaxation post-exercise dynamics and is made available on PhysioNet ...
Document Type: conference object
Language: English
Relation: info:eu-repo/grantAgreement/ScienceFundRS/Ideje/7754338/RS//; info:eu-repo/grantAgreement/MESTD/inst-2020/200017/RS//; info:eu-repo/grantAgreement/MESTD/inst-2020/200103/RS//; https://vinar.vin.bg.ac.rs/handle/123456789/15843; http://vinar.vin.bg.ac.rs/bitstream/id/44717/ICBES_145.pdf
DOI: 10.11159/icbes25.145
Availability: https://vinar.vin.bg.ac.rs/handle/123456789/15843; https://doi.org/10.11159/icbes25.145; http://vinar.vin.bg.ac.rs/bitstream/id/44717/ICBES_145.pdf
Rights: openAccess ; https://creativecommons.org/licenses/by/4.0/ ; BY
Accession Number: edsbas.C8343FEC
Database: BASE