Faculty of Computer Science and Engineering

Permanent URI for this communityhttps://repository.ukim.mk/handle/20.500.12188/5

The Faculty of Computer Science and Engineering (FCSE) within UKIM is the largest and most prestigious faculty in the field of computer science and technologies in Macedonia, and among the largest faculties in that field in the region. The FCSE teaching staff consists of 50 professors and 30 associates. These include many “best in field” personnel, such as the most referenced scientists in Macedonia and the most influential professors in the ICT industry in the Republic of Macedonia.

Browse

Search Results

Now showing 1 - 10 of 12

Some of the metrics are blocked by your
consent settings
A Method to Detect Ventricular Fibrillation in Electrocardiograms
(IEEE, 2021-09-27)
Temelkov, G.
;
The objective of this research is to create an algorithm for automatic detection of ventricular fibrillation in electrocardiogram records customized for wearable single-channel sensors. Our approach is based on observing and examining sequences of ventricular fibrillation in their frequency-domain. The research, design, and validation process, along with comprehensive annotations, is carried out on Physionet reference databases. We use a sliding window approach and apply the Fast Fourier Transform to convert the data from the time domain to the frequency domain. Our approach is based on the determination of frequency peaks, calculation of energy around the peak, and its ratio to the overall spectra. The evaluation of detection performance classification results by applying the digital signal processing algorithms with machine learning methods classify ventricular arrhythmia episodes with F1 score of 0.77 and accuracy of 0.92.
Some of the metrics are blocked by your
consent settings
TPC-H Benchmark Q3, Q6 and Q12 Sequential, OpenMP Parallel and CUDA Parallel Implementation
(IEEE, 2021-09-27)
Mitrovski, Jovan
;
Djinevski, Leonid
;
;
Arsenovski, Sime
In the last decade or two, parallel computing has solved a myriad of problems when it comes to speeding up the problem-solving process. To deliver the necessary acceleration of their product, application developers have turned to CUDA-powered GPU parallel processing. One of the many uses of the well-known NVIDIA parallel computing platform CUDA is to enhance the performance of problems whose root comes from speeding up database queries. The fundamental focus of this paper is evaluating the performance of different parallelizing techniques using the TPC-H benchmark. This is achieved by parallelizing the ‘SELECT’ queries that the benchmark offers using the Open Multi-Processing (OpenMP) API (Q3, Q6, and Q12 in particular), then parallel on the GPU using the previously mentioned CUDA platform. Lastly, comparing the GPU's performance with both sequential instruction execution and parallel instruction execution on a CPU. This implementation resulted in speedup ratios ranging from 2 to 4 when analyzing the CPU-parallel code and ranging from 10 up to 558 when analyzing the GPU - parallel code.
Some of the metrics are blocked by your
consent settings
A New ML-based AFIB Detector
(IEEE, 2021-11-23)
Tudjarski, Stojancho
;
Ignjatov, Tomislav
;
Objectives: This paper aims to develop a model for detecting Atrial Fibrillation (AFIB) in a single-channel electrocardiogram.Methodology: The applied Machine Learning methods are XGBoost and Random Forest. Training and testing split was realized by splitting the patients 70% for training and 30% for testing. Features included annotations for heartbeats, intervals between neighboring heartbeats, Shannon entropy, and Fluctuation index by designing a moving window with predefined length.Data: Standard ECG benchmarks were used for training and testing from MIT-BIH Arrhythmia [1] and MIT-BIH Atrial Fibrillation[2] datasets.Conclusion: Experiments were done with different window sizes and different hyperparameters. The best results were achieved from the 41 Beat window, with XGBoost achieving the best performance of 99% with an F1 score of 99%.
Some of the metrics are blocked by your
consent settings
Detection of Premature Heartbeats
(IEEE, 2022-05-23)
Premature heartbeats are those that appear earlier than the regular ones due to contractions not originating from the Sinus Atrial Node, out of the normal heart rhythm. Although one might think this is a trivial task to detect, the distribution of premature heartbeats in the benchmark electrocardiograms shows it is not the case. We aim at finding the optimal method to detect premature heartbeats, threshold value and context level for temporal based approach. Methodology: Several methods are specified to calculate the relation of the premature heartbeat to a set of several previous instances and conduct experiments to present which averaging method (arithmetic, geometric and harmonic mean) delivers the best solution. Then, we calculate the deviation ratio to the average of these beats that affect the prematurity condition. Particularly, the goal is to find an average AV G of previous k beat-to-beat intervals RR, such that the ratio between the difference dRR of the analyzed RR and AV G versus AV G is more than a specific threshold T hr. Data: The comprehensive MIT-BIH Arrhythmia Electrocardiogram benchmark database is used in our evaluation. The analysis is conducted on the array of beat-to-beat intervals for the heartbeats preceding the premature one. Conclusion: The results show that the optimal method is based on the arithmetic mean AM We found that the larger k, the better performance is achieved. A sufficient performance with F1 score over 85% is achieved for k = 5 and T hr = 15%.
Some of the metrics are blocked by your
consent settings
Detecting Ventricular Beats with Machine Learning Models
(IEEE, 2022-05-23)
Tudjarski, Stojancho
;
Stankovski, Aleksandar
;
This paper aims at modeling a classifier of Ventricular heartbeats by experimenting with the most advanced classic binary classifiers in different scenarios for feature engineering. Methodology: The results were acquired based on experimenting with XGBoost and Random Forest algorithms, as two of the most advanced classifiers not based on neural networks. Although the annotated ECG data sets contain records with several heartbeat classes, we focus on a model that would distinguish V from others (Non-V heartbeats). Considering that we are dealing with a highly imbalanced data set, we applied the SMOTE algorithm for data enrichment to provide a better-balanced data set for training the model. To acquire better results, we added new calculated features, with and without feature selection. For feature selection, we used the Fisher Selector algorithm. Data: We used MIT-BIH Arrhythmia benchmark database, with train/test split according to the patient-oriented splitting approach that separates the original dataset into two subsets with approximately equal sizes and distribution of heartbeat types. Conclusion: The best results are achieved with XGBoost algorithm with original feature set. We achieved precision of 91.36%, recall of 88.31% and F1 score of 89.81%. Results showed that oversampling does not provide significantly better overall model performance. Still, we would recommend this approach since in practice, when dealing with imbalanced data sets, this leads to more robust models that perform better with data outside the training and test sets, such as when the model is used in production.
Some of the metrics are blocked by your
consent settings
From Continuous ECG Signals to Extracted Features for Machine Learning Models and Arrhythmia Annotations
(IEEE, 2022-11-15)
Tudjarski, Stojancho
;
Stankovski, Aleksandar
;
;
This paper describes the process of transforming an ECG signal as a continuous stream of numbers representing measured electrical voltages between the ECG electrodes into an output indicating the existence of arrhythmia. The ECG data stream is a structured array of converted analog signal values to digital data. Although this stream uses a continuous structure of numbers within a given range that depends on the bit resolution during conversion, it is still unstructured as a representation of the appearance of arrhythmia since it does not contain information about detected arrhythmia. This paper presents how to process ECG data, detect heartbeat annotations, and calculate various parameters for tabular-based data with a fixed number of columns to be used as input into ML-based algorithms. Our use case addresses an ML algorithm to detect atrial fibrillation arrhythmia, as an irregular heart rhythm. Practically a set of numbers in the ECG samples, which do not have structured arrhythmia annotations, is transformed into structured annotations. Experiments are conducted on the well-known ECG benchmark MIT-BIH Arrhythmia Database. The input data is resampled from 360 Hz to 125 Hz signal, and a signal processing algorithm is used to detect heartbeats, extracting a fixed set of features, and systematically forwarded to the feature selection ML methodologies to obtain atrial fibrillation annotations.
Some of the metrics are blocked by your
consent settings
Sampling Rate Impact on Heart Rate Variability
(IEEE, 2022-11-15)
;
Tudjarski, Stojancho
;
Anagelevska, Ana
We set a research question to find the impact of the sampling rate in the electrocardiograms on the calculation of the heart rate variability. Our evaluation method analyzes the heart rate variability on the MITDB electrocardiogram benchmark database originally sampled to 360 Hz. Then the electrocardiogram records are resampled to 125 Hz, and a new set of heart rate variability parameters are calculated and compared to those calculated on 360 Hz. The analysis performed on time-domain heart rate variability parameters showed that the impact of sampling frequency was large for NN50 and pNN50 parameters (up to a max value of 50%, and mean of 10%), and smaller for SDNN and RMSSD (up to a max value of 10%, and an average of 2%). In conclusion, heart rate variability needs a higher sampling frequency to reveal more precise results.
Some of the metrics are blocked by your
consent settings
Effect of Bézier Interpolation on Similarity of Heartbeats in Electrocardiograms
(IEEE, 2022-11-15)
;
Tonkovikj, Lucija
;
Petrovski, Nikola
Sampling frequency and bit resolution impact the presentation of heartbeats in electrocardiograms. One of the most challenging tasks in automated medical interpretation is the determination of the heartbeat type, whether the beat was initiated by a normal sequence, or by the ventricle. The usual way to detect the beat class is to analyze the morphological features including shape, width and height of the heartbeat. In this paper, we tackle this problem by checking the similarity of the heartbeats and set a research hypothesis that Bézier interpolation can improve the correctness of the similarity check. We found that the similarity check of two heartbeats can be improved a lot by using Bézier interpolation on the samples around the detected peak.
Some of the metrics are blocked by your
consent settings
Improving Atrial Fibrillation Detection with Machine Learning Models
(IEEE, 2025-06-02)
Tudjarski, Stojancho
;
;
Madevska Bogdanova, Ana
;
Stankovski, Aleksandar
In this paper, we focus on feature engineering to detect Atrial Fibrillation by determining whether the heart rhythm has irregularities without patterns. We experiment with a broad spectrum of features derived from the duration of heartbeat-to-heartbeat intervals in the benchmark electrocardiogram databases MIT-BIH Arrhythmia Database, MIT-BIH Atrial Fibrillation Database, and Long Term AF Database. The experiments included position-based features, fluctuation indices, standard deviations, mean average values, Shannon entropy, and statistical measures, such as compressed time series data length. The research questions are to detect the most influential features that result in the best-performing model and the impact of the training dataset. Our approach is to evaluate the model on a completely different dataset from the one it was trained on. Achieved F1 scores vary between 62,32% and 85.08%. The results prove that positional features increase the model's performance.
Some of the metrics are blocked by your
consent settings
Scalability Evaluation of HPC Multi-GPU Training for ECG-based LLMs
(IEEE, 2025-06-02)
;
Petrovski, Nikola
;
Training large language models requires extensive processing, made possible by many high-performance computing resources. This study compares multi-node and multi-GPU environments for training large language models using electrocardiogram data. It provides a detailed mapping of current frameworks for distributed deep learning in multi-node and multi-GPU settings, including Horovod from Uber, DeepSpeed from Microsoft, and the built-in distributed capabilities of PyTorch and TensorFlow. We compare various multi-GPU setups for different dataset configurations, utilizing multiple HPC nodes independently and focusing on scalability, speedup, efficiency, and overhead. The analysis leverages HPC infrastructure with SLURM, Apptainer (Singularity) containers, CUDA, PyTorch, and shell scripts to support training workflows and automation. We achieved a sub-linear speedup when scaling the number of GPUs, with values of 1.6x for two GPUs and 1.9x for four GPUs.

Faculty of Computer Science and Engineering

Browse

Filters

Settings

Sort By

Results per page

Search Results