Marjan Miletić

Marjan Miletić

Auto-generated profile, based on publicly available information

Marjan Miletić

National Institute of the Republic of Serbia, University of Belgrade

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
25 March 2024 in IEEE Open Journal of Engineering in Medicine and Biology
Verified
Journal Article
Open access
VAVladimir Atanasoski
Vladimir Atanasoski
JPJovana Petrović
Jovana Petrović
LPLana Popovic-Maneski
Lana Popovic-Maneski
MMMarjan Miletić
Marjan Miletić
MBMIlos Babic
MIlos Babic
MIMarija Ivanovic
Marija Ivanovic
Goal: Clinical interpretation of an electrocardiogram (ECG) can be detrimentally affected by noise. Removal of the electromyographic (EMG) noise is particularly challenging due to its spectral overlap with the QRS complex. The existing EMG-denoising algorithms often distort signal morphology, thus obscuring diagnostically relevant information. Methods: Here, a new iterative regeneration method (IRM) for efficient EMG-noise suppression is proposed. The main hypothesis is that the temporary removal of the dominant ECG components enables extraction of the noise with the minimum alteration to the signal. The method is validated on SimEMG database of simultaneously recorded reference and noisy signals, MIT-BIH arrhythmia database and synthesized ECG signals, both with the noise from MIT Noise Stress Test Database. Results: IRM denoising and morphology-preserving performance is superior to the wavelet- and FIR-based benchmark methods. Conclusions : IRM is reliable, computationally non-intensive, fast and applicable to any number of ECG channels recorded by mobile or standard ECG devices.

3 June 2024 in IEEE Access
Verified
Journal Article
Open access
MMMarjan N. Miletić
Marjan N. Miletić
VAVladimir Atanasoski
Vladimir Atanasoski
PBPetra P. Beličev
Petra P. Beličev
GGGoran M. Gligorić
Goran M. Gligorić
URUroš M. Ralević
Uroš M. Ralević
BVBranislav Vajdić
Branislav Vajdić
Clinical outcomes of several acute conditions, including myocardial infarction (MI), the most common cause of death, can be improved by timely diagnostics based on electrocardiography (ECG). However, current diagnostic technologies include a large number of wired ECG electrodes, which require accurate placement by trained personnel. The ideal ECG device would be suitable for self-measurement, i.e., would have a small number of electrodes, be mobile or portable, and provide an accurate diagnosis. However, these aims have not been met using the same device. A recently developed handheld ECG device with three quasi-orthogonal leads opened the door for mobile assessment of the three-dimensional cardiac vector by self-measurement. We hypothesize that the information provided is sufficient for accurate reconstruction of the 12-lead ECG. We propose a reconstruction algorithm based on the segment-by-segment 4-matrix (4M) transformation applied to the P wave, QRS complex, ST segment, and T wave. The accuracy of the 4M method was tested using data obtained from 64 healthy volunteers. The 4M method reconstructed the standard 12-lead ECG with a 0.96 mean cross-correlation for all leads and provided meaningful clinical results. A back-to-back comparative study demonstrated the superiority of the proposed method over the traditional EASI method. In addition, the results provide evidence of the capability of the 3-lead 4M technology to accurately reconstruct the full cardiac vector from a single measurement, which distinguishes it from competition. Although further clinical investigation is necessary, wireless operation and high accuracy make the proposed method potentially suitable for remote monitoring and self-assessment.

7 November 2023 in IEEE Open Journal of Engineering in Medicine and Biology
Verified
Journal Article
Open access
VAVladimir Atanasoski
Vladimir Atanasoski
JPJovana Petrovic
Jovana Petrovic
LPLana Popovic-Maneski
Lana Popovic-Maneski
MMMarjan Miletić
Marjan Miletić
MBMIlos Babic
MIlos Babic
MIMarija Ivanovic
Marija Ivanovic
Goal: Noise on recorded electrocardiographic (ECG) signals may affect their clinical interpretation. Electromyographic (EMG) noise spectrally coincides with the QRS complex, which makes its removal particularly challenging. The problem of evaluating the noise-removal techniques has commonly been approached by algorithm testing on the contaminated ECG signals constructed ad hoc as an additive mixture of a noise-free ECG signal and noise. Consequently, there is an absence of a unique/standard database for testing and comparing different denoising methods. We present a SimEMG database recorded by a novel acquisition method that allows for direct recording of the genuine EMG-noise-free and -contaminated ECG signals. The database is available as open source.

1 January 2026 in Chaos: An Interdisciplinary Journal of Nonlinear Science
Verified
Journal Article
Closed access
AMAleksandra Maluckov
Aleksandra Maluckov
DSDanka B. Stojanović
Danka B. Stojanović
MMM. Miletić
M. Miletić
MIMarija Ivanovic
Marija Ivanovic
LHLj. Hadžievski
Lj. Hadžievski
JPJ. Petrović
J. Petrović
We investigate the recovery dynamics of healthy cardiac activity after physical exertion using multimodal biosignals recorded with a polycardiograph. Multifractal features derived from the singularity spectrum capture the scale-invariant properties of cardiovascular regulation. Five supervised classification algorithms—Logistic Regression (LogReg), Support Vector Machine with radial basis function kernel, k-Nearest Neighbors, Decision Tree, and Random Forest—were evaluated to distinguish recovery states in a small, imbalanced dataset. Our results show that multifractal analysis, combined with multimodal sensing, yields reliable features for characterizing recovery and points toward nonlinear diagnostic methods for heart conditions.

31 July 2024
Verified
Conference Paper
Closed access
VAVladimir Atanasoski
Vladimir Atanasoski
LPLana Popovic-Maneski
Lana Popovic-Maneski
ISInstitute of Technical Sciences of the SASA
Institute of Technical Sciences of the SASA
MIMarija Ivanovic
Marija Ivanovic
MMMarjan Miletić
Marjan Miletić
ANAleksandra Nikolic
Aleksandra Nikolic
ECG measurements at rest are susceptible to different kinds of noise which may affect clinical interpretation. Unlike the low-frequency body movements or spectrally well-defined power line interference, the broadband electromyographic (EMG) noise spectrally coincides with the QRS complex, which makes its removal particularly challenging. While a number of algorithms have been developed, their evaluation and comparison have long remained challenging due to the lack of a database with EMG-contaminated and genuine noise-free signals. We have recently published a SimEMG database of the simultaneously recorded EMG-noise-free and -contaminated ECG signals. In this paper, we examine the capability of SimEMG to become a reference test set for different denoising algorithms. We show that the SimEMG is useful for the assessment of the signal morphology distortions leading to potentially misleading changes in the diagnostic markers. We examine the ST segment – a marker of acute myocardial infarction, and R, Pmax, and Tmax points– the markers of arrhythmias. The tests are performed with the adaptive wavelet Wiener filter, wavelet transform (WT), finite impulse response (FIR) filter, and iterative regenerative method (IRM). The analysis also reveals the residual noise as a minor limitation of the database and acquisition method. Possible improvements are discussed.

9 June 2025
Verified
Conference Paper
Closed access
MMMarjan Miletić
Marjan Miletić
VAVladimir Atanasoski
Vladimir Atanasoski
PBPetra P. Beličev
Petra P. Beličev
GGGoran Gligorić
Goran Gligorić
URUroš Ralević
Uroš Ralević
BVBranislav Vajdić
Branislav Vajdić
Mobile ECG devices effectively detect arrhythmias but often miss acute conditions like heart attacks due to the limited data from just 1-2 leads. We have recently proposed a novel method of ECG synthesis for synthesize 12-lead ECG from 3 quasiorthogonal ECG leads using a 4-matrix (4M) transformation. The approach employs four personalized matrices to reconstruct the P wave, QRS complex, ST segment, and T wave, calibrated from non-simultaneous 3-lead and 12-lead ECG measurements. Here we provide a detailed account on the method testing on the recordings obtained from 64 healthy volunteers. The method outperforms the EASI-based synthesis, achieving a mean cross-correlation (CC) of 0.96 versus 0.89 for EASI. Morphology and cardiac angle agreement, assessed via Pearson R-correlation, is also superior. The 4 M method enables reliable diagnostic interpretation, making it suitable for converting mobile 3-lead ECG device data into accurate synthesized 12-lead recordings.