| Volume 29, Issue 4 | |
| Comparative Analysis of Technical Parameters of Acupuncture-like Electrostimulation Device | 249-256 |
| Gadir Gafarov | |
| doi: 10.7546/ijba.2025.29.4.001006 | |
| [ +/- how to cite ][ +/- abstract ][ full text ] | |
| Gadir Gafarov (2025) Comparative Analysis of Technical Parameters of Acupuncture-like Electrostimulation Device, Int J Bioautomation, 29 (4), 249-256, doi: 10.7546/ijba.2025.29.4.001006 | |
| Abstract: The comparative analysis of technical parameters of acupuncture-like electrostimulation devices provides crucial insights into their performance, efficacy, and usability in therapeutic applications. This study evaluates a selection of commercially available electrostimulation devices, focusing on key parameters such as output frequency, pulse duration, current intensity, waveform characteristics, and safety features. A comparative analysis of the output signal frequencies of an acupuncture-like electrostimulator and commercial electrostimulator devices was designed and performed. The error percentage in the output signal frequency of the designed electrostimulator was determined. It was found that the designed stimulator is superior in terms of signal stability compared to existing electrostimulator devices.
Keywords: Acupuncture-like electrostimulation, Therapeutic waveform characteristics, Therapeutic applications, Transcutaneous electrical nerve stimulation (TENS), Acupuncture | |
| A Dual-stage Approach for Chronic Kidney Disease Detection Using Shapley Additive Explanations and an Evolutionary Light Gradient Boosting Machine | 257-284 |
| Moumita Pramanik, Ranjit Panigrahi, Bidita Khandelwal, Joseph Bamidele Awotunde, Akash Kumar Bhoi | |
| doi: 10.7546/ijba.2025.29.4.001017 | |
| [ +/- how to cite ][ +/- abstract ][ full text ] | |
| Moumita Pramanik, Ranjit Panigrahi, Bidita Khandelwal, Joseph Bamidele Awotunde, Akash Kumar Bhoi (2025) A Dual-stage Approach for Chronic Kidney Disease Detection Using Shapley Additive Explanations and an Evolutionary Light Gradient Boosting Machine, Int J Bioautomation, 29 (4), 257-284, doi: 10.7546/ijba.2025.29.4.001017 | |
| Abstract: This article discusses significant advancements in diagnosing chronic kidney disease (CKD) using state-of-the-art machine learning methods. It employs two CKD detection models: the gradient boosting decision tree (GBDT) model and dropout additive regression tree (DART) model. The models are used for CKD detection within the light gradient boosting machine (LightGBM) framework. This article also describes a new dual-stage feature selection strategy to improve model inputs by selecting relevant features while maintaining transparency. The LightGBM feature importance score was used to identify critical features of CKD patients during the feature selection phase. Additionally, the SHapley Additive exPlanations (SHAP) method is utilised to assess the significance of individual attributes, making the model predictions easier to understand. The experimental evaluation utilised a dataset containing 24 features related to CKD. The developed GBDT and DART models demonstrated high accuracy, sensitivity, and specificity levels. The GBDT model exhibited a sensitivity of 99.20%, a specificity of 100%, and an accuracy of 99.50%. Similarly, the GBDT-based model achieved a precision of 98.80%, a sensitivity of 99.20%, and a specificity of 100%.
Keywords: Chronic kidney disease, Machine learning, Gradient boosting decision trees, Light gradient boosting machine framework, SHapley Additive exPlanations | |
| Machine Learning Approaches to Laryngeal Pathologies Detection and Classification: A Comprehensive Literature Review | 285-314 |
| Hassan Ezzahori, Abdelkrim Hammimou, Abdelghani Boudaoud, Mounaim Aqil | |
| doi: 10.7546/ijba.2025.29.4.001071 | |
| [ +/- how to cite ][ +/- abstract ][ full text ] | |
| Hassan Ezzahori, Abdelkrim Hammimou, Abdelghani Boudaoud, Mounaim Aqil (2025) Machine Learning Approaches to Laryngeal Pathologies Detection and Classification: A Comprehensive Literature Review, Int J Bioautomation, 29 (4), 285-314, doi: 10.7546/ijba.2025.29.4.001071 | |
| Abstract: Voice alterations are the most frequent early sign of laryngeal pathologies. Implementing voice screening protocols could help to detect early laryngeal diseases, such as cancer. This field is increasingly turning to machine learning (ML) approaches for diagnostic purposes, using analysis of vocal patterns and speech characteristics to identify these disorders. This review aims to synthesize and evaluate the literature on laryngeal disease detection and classification using ML. A comprehensive search of five major multidisciplinary and specialized databases was conducted to identify articles published between 2015-2024, yielding 102 relevant studies. Data extraction and analysis were conducted using the “preferred reporting items for systematic reviews and meta-analyses” system. The included studies utilize deep learning or ML algorithms to analyze the speech signal. The review reveals that the Saarbrücken voice database remains the most coveted by the researchers, as it was used in 53% of the studies. It shows that mel-frequency cepstral coefficients are the most commonly used features, appearing in 54% of included studies, alongside the support vector machine algorithm, which is the most commonly used classifier (50% of the studies). The review demonstrates that traditional ML techniques are constantly being overtaken by deep learning ones. This review serves as a roadmap for future research, guiding the development of ML and deep learning-based algorithms for laryngeal disease detection.
Keywords: Voice disorders, Neural network, Machine learning, Deep learning, Laryngeal disease | |
| Processing and Analysis of EMG Signals from MYO Armband for Upper-Limb Prosthesis Control | 315-326 |
| Yoto Yotov, Emil Petrov, Velislava Lyubenova | |
| doi: 10.7546/ijba.2025.29.4.001113 | |
| [ +/- how to cite ][ +/- abstract ][ full text ] | |
| Yoto Yotov, Emil Petrov, Velislava Lyubenova (2025) Processing and Analysis of EMG Signals from MYO Armband for Upper-Limb Prosthesis Control, Int J Bioautomation, 29 (4), 315-326, doi: 10.7546/ijba.2025.29.4.001113 | |
| Abstract: This paper presents the development of an electromyographic (EMG) signal processing system for controlling upper limb prostheses using data from the MYO Armband, which includes eight EMG sensors. The system processes raw EMG signals through filtering, frequency analysis, and other enhancement techniques to improve signal quality. An interactive MATLAB interface, built on the Myo SDK MATLAB MEX Wrapper, enables real-time visualization and application of various filters. A comprehensive comparison of filtering methods assesses their influence on signal reliability and performance. Quantitative results indicate that the Power Grip gesture produces the highest EMG activation, while the Extended Index Finger shows lower muscle engagement, highlighting distinct activation patterns. Heat map visualizations reveal spatial activation differences across sensors, essential for designing effective gesture classifiers. The developed platform enhances noise robustness and improves accuracy in interpreting motor commands. Despite hardware limitations, the system demonstrates the feasibility of adaptive prosthesis control and suggests integration with hybrid methods, such as voice control, to further enhance functionality and user experience.
Keywords: Myoelectric signals, Signal filtering, Surface electromyography, MATLAB interface, MYO Armband | |
| Enhancing Model Accuracy: A Parameter Optimization Strategy Based on the Dream Optimization Algorithm | 327-339 |
| Emanuela Yaneva, Olympia Roeva | |
| doi: 10.7546/ijba.2025.29.4.001116 | |
| [ +/- how to cite ][ +/- abstract ][ full text ] | |
| Emanuela Yaneva, Olympia Roeva (2025) Enhancing Model Accuracy: A Parameter Optimization Strategy Based on the Dream Optimization Algorithm , Int J Bioautomation, 29 (4), 327-339, doi: 10.7546/ijba.2025.29.4.001116 | |
| Abstract: Reliable parameter identification is essential for the development and predictive use of non-linear bioprocess models. This study evaluates the recently proposed Dream Optimization Algorithm (DOA), a human-inspired metaheuristic based on memory retention, partial forgetting, and dream-sharing mechanisms, for the identification of kinetic parameters in an Escherichia coli fed-batch cultivation model. The algorithm’s performance is assessed using experimental cultivation data and compared with three widely employed metaheuristics: the genetic algorithm (GA), simulated annealing (SA), and the crow search algorithm (CSA). Results demonstrate that DOA achieves the lowest objective function value, the best mean performance across 30 independent runs, and substantially reduced computational time compared to SA and CSA. The model dynamics generated using DOA-identified parameters show excellent agreement with experimental biomass and substrate measurements, even in the presence of significant noise in the substrate data. These findings highlight the high accuracy, robustness, and computational efficiency of DOA, confirming its strong potential as an effective tool for bioprocess model parameter estimation and broader non-linear optimization tasks.
Keywords: Dream optimization algorithm, Metaheuristic, Escherichia coli, Model parameter identification |
© 2025, BAS, Institute of Biophysics and Biomedical Engineering