TY  - CONF
AU  - Tadić, Predrag
AU  - Petrović, Jovana
AU  - Đorđević, Natalija
AU  - Ivanović, Marija
AU  - Lazović, Aleksandar
AU  - Vukčević, Vladan
AU  - Ristić, Arsen
AU  - Hadžievski, Ljupčo
PY  - 2023
UR  - https://vinar.vin.bg.ac.rs/handle/123456789/13044
AB  - We present our progress on the “Multi-SENSor SysteM and ARTificial intelligence in service of heart failure diagnosis (SensSmart)” project, which was introduced at the last year’s edition of the Workshop [1]. The goal of the SensSmart project is to enable early diagnosis of heart failure, through the development of: 1) a multi-sensor polycardiograph apparatus (PCG) that produces simultaneous acquisition of the subject’s electrocardiogram (ECG), photoplethysmogram (PPG), heart sounds, and heart movements, and 2) AI-assisted analysis of the acquired signals. This presentation is going to focus on the acquisition and processing of PPG signals. PPG is obtained by using a pulse oximeter which illuminates the skin and measures the changes in light absorption, thereby enabling the detection of blood volume changes in the vessels. Our PCG apparatus measures the blood flow through the brachial, radial, and carotid arteries. During each heartbeat, the generated waveform typically exhibits several characteristic points [2]. The magnitudes and time distances between these points are useful indicators of many cardiac conditions, including heart failure [3]. However, the inter-patient variability of the PPG waveform makes it challenging to derive simple rule-based diagnostic procedures. This has led many researchers to turn to statistical or machine learning methods for processing of PPG signals [4].  In this presentation, we give an overview of AI-based signal processing methods for PPG, and present some preliminary results and challenges in extracting features from real-world signals obtained using our PCG.
PB  - Belgrade : Institute of Physics
C3  - 16th Photonics Workshop : Book of abstracts
T1  - Photoplethysmogram as a source of biomarkers  for AI-based diagnosis of heart failure
SP  - 24
EP  - 24
UR  - https://hdl.handle.net/21.15107/rcub_vinar_13044
ER  - 

@conference{
author = "Tadić, Predrag and Petrović, Jovana and Đorđević, Natalija and Ivanović, Marija and Lazović, Aleksandar and Vukčević, Vladan and Ristić, Arsen and Hadžievski, Ljupčo",
year = "2023",
abstract = "We present our progress on the “Multi-SENSor SysteM and ARTificial intelligence in service of heart failure diagnosis (SensSmart)” project, which was introduced at the last year’s edition of the Workshop [1]. The goal of the SensSmart project is to enable early diagnosis of heart failure, through the development of: 1) a multi-sensor polycardiograph apparatus (PCG) that produces simultaneous acquisition of the subject’s electrocardiogram (ECG), photoplethysmogram (PPG), heart sounds, and heart movements, and 2) AI-assisted analysis of the acquired signals. This presentation is going to focus on the acquisition and processing of PPG signals. PPG is obtained by using a pulse oximeter which illuminates the skin and measures the changes in light absorption, thereby enabling the detection of blood volume changes in the vessels. Our PCG apparatus measures the blood flow through the brachial, radial, and carotid arteries. During each heartbeat, the generated waveform typically exhibits several characteristic points [2]. The magnitudes and time distances between these points are useful indicators of many cardiac conditions, including heart failure [3]. However, the inter-patient variability of the PPG waveform makes it challenging to derive simple rule-based diagnostic procedures. This has led many researchers to turn to statistical or machine learning methods for processing of PPG signals [4].  In this presentation, we give an overview of AI-based signal processing methods for PPG, and present some preliminary results and challenges in extracting features from real-world signals obtained using our PCG.",
publisher = "Belgrade : Institute of Physics",
journal = "16th Photonics Workshop : Book of abstracts",
title = "Photoplethysmogram as a source of biomarkers  for AI-based diagnosis of heart failure",
pages = "24-24",
url = "https://hdl.handle.net/21.15107/rcub_vinar_13044"
}

Tadić, P., Petrović, J., Đorđević, N., Ivanović, M., Lazović, A., Vukčević, V., Ristić, A.,& Hadžievski, L.. (2023). Photoplethysmogram as a source of biomarkers  for AI-based diagnosis of heart failure. in 16th Photonics Workshop : Book of abstracts
Belgrade : Institute of Physics., 24-24.
https://hdl.handle.net/21.15107/rcub_vinar_13044

Tadić P, Petrović J, Đorđević N, Ivanović M, Lazović A, Vukčević V, Ristić A, Hadžievski L. Photoplethysmogram as a source of biomarkers  for AI-based diagnosis of heart failure. in 16th Photonics Workshop : Book of abstracts. 2023;:24-24.
https://hdl.handle.net/21.15107/rcub_vinar_13044 .

Tadić, Predrag, Petrović, Jovana, Đorđević, Natalija, Ivanović, Marija, Lazović, Aleksandar, Vukčević, Vladan, Ristić, Arsen, Hadžievski, Ljupčo, "Photoplethysmogram as a source of biomarkers  for AI-based diagnosis of heart failure" in 16th Photonics Workshop : Book of abstracts (2023):24-24,
https://hdl.handle.net/21.15107/rcub_vinar_13044 .

TY  - JOUR
AU  - Shvilkin, Alexei
AU  - Vukajlović, Dejan
AU  - Bojović, Boško P.
AU  - Hadžievski, Ljupčo
AU  - Vajdic, Branislav
AU  - Atanasoski, Vladimir
AU  - Miletić, Marjan
AU  - Zimetbaum, Peter J.
AU  - Gibson, C. Michael
AU  - Vukčević, Vladan
PY  - 2023
UR  - https://vinar.vin.bg.ac.rs/handle/123456789/12393
AB  - Background Early coronary occlusion detection by portable personal device with limited number of electrocardiographic (ECG) leads might shorten symptom-to-balloon time in acute coronary syndromes.  Objectives The purpose of this study was to compare the accuracy of coronary occlusion detection using vectorcardgiographic analysis of a near-orthogonal 3-lead ECG configuration suitable for credit card-size personal device integration with automated and human 12 lead ECG interpretation.  Methods The 12-lead ECGs with 3 additional leads (“abc”) using 2 arm and 2 left parasternal electrodes were recorded in 66 patients undergoing percutaneous coronary intervention prior to (“baseline”, n = 66), immediately before (“preinflation”, n = 66), and after 90-second balloon coronary occlusion (“inflation”, n = 120). Performance of computer-measured ST-segment shift on vectorcardgiographic loops constructed from “abc” and 12 leads, standard 12-lead ECG, and consensus human interpretation in coronary occlusion detection were compared in “comparative” and “spot” modes (with/without reference to “baseline”) using areas under ROC curves (AUC), reliability, and sensitivity/specificity analysis.  Results Comparative “abc”-derived ST-segment shift was similar to two 12-lead methods (vector/traditional) in detecting balloon coronary occlusion (AUC = 0.95, 0.96, and 0.97, respectively, P = NS). Spot “abc” and 12-lead measurements (AUC = 0.72, 0.77, 0.68, respectively, P = NS) demonstrated poorer performance (P < 0.01 vs comparative measurements). Reliability analysis demonstrated comparative automated measurements in “good” agreement with reference (preinflation/inflation), while comparative human interpretation was in “moderate” range. Spot automated and human reading showed “poor” agreement.  Conclusions Vectorcardiographic ST-segment analysis using baseline comparison of 3-lead ECG system suitable for credit card-size personal device integration is similar to established 12-lead ECG methods in detecting balloon coronary occlusion.
T2  - JACC: Advances
T1  - Coronary Artery Occlusion Detection Using 3-Lead ECG System Suitable for Credit Card-Size Personal Device Integration
VL  - 2
IS  - 6
SP  - 100454
DO  - 10.1016/j.jacadv.2023.100454
ER  - 

@article{
author = "Shvilkin, Alexei and Vukajlović, Dejan and Bojović, Boško P. and Hadžievski, Ljupčo and Vajdic, Branislav and Atanasoski, Vladimir and Miletić, Marjan and Zimetbaum, Peter J. and Gibson, C. Michael and Vukčević, Vladan",
year = "2023",
abstract = "Background Early coronary occlusion detection by portable personal device with limited number of electrocardiographic (ECG) leads might shorten symptom-to-balloon time in acute coronary syndromes.  Objectives The purpose of this study was to compare the accuracy of coronary occlusion detection using vectorcardgiographic analysis of a near-orthogonal 3-lead ECG configuration suitable for credit card-size personal device integration with automated and human 12 lead ECG interpretation.  Methods The 12-lead ECGs with 3 additional leads (“abc”) using 2 arm and 2 left parasternal electrodes were recorded in 66 patients undergoing percutaneous coronary intervention prior to (“baseline”, n = 66), immediately before (“preinflation”, n = 66), and after 90-second balloon coronary occlusion (“inflation”, n = 120). Performance of computer-measured ST-segment shift on vectorcardgiographic loops constructed from “abc” and 12 leads, standard 12-lead ECG, and consensus human interpretation in coronary occlusion detection were compared in “comparative” and “spot” modes (with/without reference to “baseline”) using areas under ROC curves (AUC), reliability, and sensitivity/specificity analysis.  Results Comparative “abc”-derived ST-segment shift was similar to two 12-lead methods (vector/traditional) in detecting balloon coronary occlusion (AUC = 0.95, 0.96, and 0.97, respectively, P = NS). Spot “abc” and 12-lead measurements (AUC = 0.72, 0.77, 0.68, respectively, P = NS) demonstrated poorer performance (P < 0.01 vs comparative measurements). Reliability analysis demonstrated comparative automated measurements in “good” agreement with reference (preinflation/inflation), while comparative human interpretation was in “moderate” range. Spot automated and human reading showed “poor” agreement.  Conclusions Vectorcardiographic ST-segment analysis using baseline comparison of 3-lead ECG system suitable for credit card-size personal device integration is similar to established 12-lead ECG methods in detecting balloon coronary occlusion.",
journal = "JACC: Advances",
title = "Coronary Artery Occlusion Detection Using 3-Lead ECG System Suitable for Credit Card-Size Personal Device Integration",
volume = "2",
number = "6",
pages = "100454",
doi = "10.1016/j.jacadv.2023.100454"
}

Shvilkin, A., Vukajlović, D., Bojović, B. P., Hadžievski, L., Vajdic, B., Atanasoski, V., Miletić, M., Zimetbaum, P. J., Gibson, C. M.,& Vukčević, V.. (2023). Coronary Artery Occlusion Detection Using 3-Lead ECG System Suitable for Credit Card-Size Personal Device Integration. in JACC: Advances, 2(6), 100454.
https://doi.org/10.1016/j.jacadv.2023.100454

Shvilkin A, Vukajlović D, Bojović BP, Hadžievski L, Vajdic B, Atanasoski V, Miletić M, Zimetbaum PJ, Gibson CM, Vukčević V. Coronary Artery Occlusion Detection Using 3-Lead ECG System Suitable for Credit Card-Size Personal Device Integration. in JACC: Advances. 2023;2(6):100454.
doi:10.1016/j.jacadv.2023.100454 .

Shvilkin, Alexei, Vukajlović, Dejan, Bojović, Boško P., Hadžievski, Ljupčo, Vajdic, Branislav, Atanasoski, Vladimir, Miletić, Marjan, Zimetbaum, Peter J., Gibson, C. Michael, Vukčević, Vladan, "Coronary Artery Occlusion Detection Using 3-Lead ECG System Suitable for Credit Card-Size Personal Device Integration" in JACC: Advances, 2, no. 6 (2023):100454,
https://doi.org/10.1016/j.jacadv.2023.100454 . .

TY  - JOUR
AU  - Ivanović, Marija D.
AU  - Hannink, Julius
AU  - Ring, Matthias
AU  - Baronio, Fabio
AU  - Vukčević, Vladan
AU  - Hadžievski, Ljupčo
AU  - Eskofier, Bjoern
PY  - 2020
UR  - https://vinar.vin.bg.ac.rs/handle/123456789/9678
AB  - Optimizing timing of defibrillation by evaluating the likelihood of a successful outcome could significantly enhance resuscitation. Previous studies employed conventional machine learning approaches and hand-crafted features to address this issue, but none have achieved superior performance to be widely accepted. This study proposes a novel approach in which predictive features are automatically learned.MethodsA raw 4s VF episode immediately prior to first defibrillation shock was feed to a 3-stage CNN feature extractor. Each stage was composed of 4 components: convolution, rectified linear unit activation, dropout and max-pooling. At the end of feature extractor, the feature map was flattened and connected to a fully connected multi-layer perceptron for classification. For model evaluation, a 10 fold cross-validation was employed. To balance classes, SMOTE oversampling method has been applied to minority class.ResultsThe obtained results show that the proposed model is highly accurate in predicting defibrillation outcome (Acc = 93.6 %). Since recommendations on classifiers suggest at least 50 % specificity and 95 % sensitivity as safe and useful predictors for defibrillation decision, the reported sensitivity of 98.8 % and specificity of 88.2 %, with the analysis speed of 3 ms/input signal, indicate that the proposed model possesses a good prospective to be implemented in automated external defibrillators.ConclusionsThe learned features demonstrate superiority over hand-crafted ones when performed on the same dataset. This approach benefits from being fully automatic by fusing feature extraction, selection and classification into a single learning model. It provides a superior strategy that can be used as a tool to guide treatment of OHCA patients in bringing optimal decision of precedence treatment. Furthermore, for encouraging replicability, the dataset has been made publicly available to the research community.
T2  - Artificial Intelligence in Medicine
T1  - Predicting defibrillation success in out-of-hospital cardiac arrested patients: Moving beyond feature design
VL  - 110
SP  - 101963
DO  - 10.1016/j.artmed.2020.101963
ER  - 

@article{
author = "Ivanović, Marija D. and Hannink, Julius and Ring, Matthias and Baronio, Fabio and Vukčević, Vladan and Hadžievski, Ljupčo and Eskofier, Bjoern",
year = "2020",
abstract = "Optimizing timing of defibrillation by evaluating the likelihood of a successful outcome could significantly enhance resuscitation. Previous studies employed conventional machine learning approaches and hand-crafted features to address this issue, but none have achieved superior performance to be widely accepted. This study proposes a novel approach in which predictive features are automatically learned.MethodsA raw 4s VF episode immediately prior to first defibrillation shock was feed to a 3-stage CNN feature extractor. Each stage was composed of 4 components: convolution, rectified linear unit activation, dropout and max-pooling. At the end of feature extractor, the feature map was flattened and connected to a fully connected multi-layer perceptron for classification. For model evaluation, a 10 fold cross-validation was employed. To balance classes, SMOTE oversampling method has been applied to minority class.ResultsThe obtained results show that the proposed model is highly accurate in predicting defibrillation outcome (Acc = 93.6 %). Since recommendations on classifiers suggest at least 50 % specificity and 95 % sensitivity as safe and useful predictors for defibrillation decision, the reported sensitivity of 98.8 % and specificity of 88.2 %, with the analysis speed of 3 ms/input signal, indicate that the proposed model possesses a good prospective to be implemented in automated external defibrillators.ConclusionsThe learned features demonstrate superiority over hand-crafted ones when performed on the same dataset. This approach benefits from being fully automatic by fusing feature extraction, selection and classification into a single learning model. It provides a superior strategy that can be used as a tool to guide treatment of OHCA patients in bringing optimal decision of precedence treatment. Furthermore, for encouraging replicability, the dataset has been made publicly available to the research community.",
journal = "Artificial Intelligence in Medicine",
title = "Predicting defibrillation success in out-of-hospital cardiac arrested patients: Moving beyond feature design",
volume = "110",
pages = "101963",
doi = "10.1016/j.artmed.2020.101963"
}

Ivanović, M. D., Hannink, J., Ring, M., Baronio, F., Vukčević, V., Hadžievski, L.,& Eskofier, B.. (2020). Predicting defibrillation success in out-of-hospital cardiac arrested patients: Moving beyond feature design. in Artificial Intelligence in Medicine, 110, 101963.
https://doi.org/10.1016/j.artmed.2020.101963

Ivanović MD, Hannink J, Ring M, Baronio F, Vukčević V, Hadžievski L, Eskofier B. Predicting defibrillation success in out-of-hospital cardiac arrested patients: Moving beyond feature design. in Artificial Intelligence in Medicine. 2020;110:101963.
doi:10.1016/j.artmed.2020.101963 .

Ivanović, Marija D., Hannink, Julius, Ring, Matthias, Baronio, Fabio, Vukčević, Vladan, Hadžievski, Ljupčo, Eskofier, Bjoern, "Predicting defibrillation success in out-of-hospital cardiac arrested patients: Moving beyond feature design" in Artificial Intelligence in Medicine, 110 (2020):101963,
https://doi.org/10.1016/j.artmed.2020.101963 . .

TY  - JOUR
AU  - Ivanović, Marija D.
AU  - Ring, Matthias
AU  - Baronio, Fabio
AU  - Calza, Stefano
AU  - Vukčević, Vladan
AU  - Hadžievski, Ljupčo
AU  - Maluckov, Aleksandra
AU  - Eskofier, Bjoern Michael
PY  - 2018
UR  - http://stacks.iop.org/2057-1976/5/i=1/a=015012?key=crossref.179380a6d1fbac3633b726787a95feb5
UR  - https://vinar.vin.bg.ac.rs/handle/123456789/8095
AB  - Objective: Algorithms to predict shock outcome based on ventricular fibrillation (VF) waveform features are potentially useful tool to optimize defibrillation strategy (immediate defibrillation versus cardiopulmonary resuscitation). Researchers have investigated numerous predictive features and classification methods using single VF feature and/or their combinations, however reported predictabilities are not consistent. The purpose of this study was to validate whether combining VF features can enhance the prediction accuracy in comparison to single feature. Approach: The analysis was performed in 3 stages: feature extraction, preprocessing and feature selection and classification. Twenty eight predictive features were calculated on 4s episode of the pre-shock VF signal. The preprocessing included instances normalization and oversampling. Seven machine learning algorithms were employed for selecting the best performin single feature and combination of features using wrapper method: Logistic Regression (LR), Naïve-Bayes (NB), Decision tree (C4.5), AdaBoost.M1 (AB), Support Vector Machine (SVM), Nearest Neighbour (NN) and Random Forest (RF). Evaluation of the algorithms was performed by nested 10 fold cross-validation procedure. Main results: A total of 251 unbalanced first shocks (195 unsuccessful and 56 successful) were oversampled to 195 instances in each class. Performance metric based on average accuracy of feature combination has shown that LR and NB exhibit no improvement, C4.5 and AB an improvement not greater than 1% and SVM, NN and RF an improvement greater than 5% in predicting defibrillation outcome in comparison to the best single feature. Significance: By performing wrapper method to select best performing feature combination the non-linear machine learning strategies (SVM, NN, RF) can improve defibrillation prediction performance. © 2018 IOP Publishing Ltd.
T2  - Biomedical Physics and Engineering Express
T1  - ECG derived feature combination versus single feature in predicting defibrillation success in out-of-hospital cardiac arrested patients
VL  - 5
IS  - 1
SP  - 015012
DO  - 10.1088/2057-1976/aaebec
ER  - 

@article{
author = "Ivanović, Marija D. and Ring, Matthias and Baronio, Fabio and Calza, Stefano and Vukčević, Vladan and Hadžievski, Ljupčo and Maluckov, Aleksandra and Eskofier, Bjoern Michael",
year = "2018",
abstract = "Objective: Algorithms to predict shock outcome based on ventricular fibrillation (VF) waveform features are potentially useful tool to optimize defibrillation strategy (immediate defibrillation versus cardiopulmonary resuscitation). Researchers have investigated numerous predictive features and classification methods using single VF feature and/or their combinations, however reported predictabilities are not consistent. The purpose of this study was to validate whether combining VF features can enhance the prediction accuracy in comparison to single feature. Approach: The analysis was performed in 3 stages: feature extraction, preprocessing and feature selection and classification. Twenty eight predictive features were calculated on 4s episode of the pre-shock VF signal. The preprocessing included instances normalization and oversampling. Seven machine learning algorithms were employed for selecting the best performin single feature and combination of features using wrapper method: Logistic Regression (LR), Naïve-Bayes (NB), Decision tree (C4.5), AdaBoost.M1 (AB), Support Vector Machine (SVM), Nearest Neighbour (NN) and Random Forest (RF). Evaluation of the algorithms was performed by nested 10 fold cross-validation procedure. Main results: A total of 251 unbalanced first shocks (195 unsuccessful and 56 successful) were oversampled to 195 instances in each class. Performance metric based on average accuracy of feature combination has shown that LR and NB exhibit no improvement, C4.5 and AB an improvement not greater than 1% and SVM, NN and RF an improvement greater than 5% in predicting defibrillation outcome in comparison to the best single feature. Significance: By performing wrapper method to select best performing feature combination the non-linear machine learning strategies (SVM, NN, RF) can improve defibrillation prediction performance. © 2018 IOP Publishing Ltd.",
journal = "Biomedical Physics and Engineering Express",
title = "ECG derived feature combination versus single feature in predicting defibrillation success in out-of-hospital cardiac arrested patients",
volume = "5",
number = "1",
pages = "015012",
doi = "10.1088/2057-1976/aaebec"
}

Ivanović, M. D., Ring, M., Baronio, F., Calza, S., Vukčević, V., Hadžievski, L., Maluckov, A.,& Eskofier, B. M.. (2018). ECG derived feature combination versus single feature in predicting defibrillation success in out-of-hospital cardiac arrested patients. in Biomedical Physics and Engineering Express, 5(1), 015012.
https://doi.org/10.1088/2057-1976/aaebec

Ivanović MD, Ring M, Baronio F, Calza S, Vukčević V, Hadžievski L, Maluckov A, Eskofier BM. ECG derived feature combination versus single feature in predicting defibrillation success in out-of-hospital cardiac arrested patients. in Biomedical Physics and Engineering Express. 2018;5(1):015012.
doi:10.1088/2057-1976/aaebec .

Ivanović, Marija D., Ring, Matthias, Baronio, Fabio, Calza, Stefano, Vukčević, Vladan, Hadžievski, Ljupčo, Maluckov, Aleksandra, Eskofier, Bjoern Michael, "ECG derived feature combination versus single feature in predicting defibrillation success in out-of-hospital cardiac arrested patients" in Biomedical Physics and Engineering Express, 5, no. 1 (2018):015012,
https://doi.org/10.1088/2057-1976/aaebec . .