BACKGROUND After 25 years of implementing the Medical Devices Directive (MDD), in 2017, the new Medical Devices Regulation (MDR) came into force, establishing stricter requirements for post-market surveillance of the safety and performance of medical devices (MD). For electrocardiogram (ECG) devices, which are crucial for monitoring cardiac activities, these requirements are essential to ensure the reliability and accuracy of diagnosing cardiac conditions and timely treatment. OBJECTIVE This study aims to enhance post-market surveillance of ECG devices by leveraging Machine Learning (ML) algorithms to predict the operational status of these devices. Specifically, the research focuses on classifying the success or failure of ECG device operations based on performance and safety parameters. The ultimate goal is to improve the management strategies of ECG devices in healthcare institutions, ensuring optimal functionality and increasing the reliability of diagnostic procedures. METHOD During the inspection process of ECG devices conducted by an accredited laboratory in accordance with ISO 17020 standard in numerous healthcare institutions in Bosnia and Herzegovina, a total of 5577 samples were collected. Various machine learning algorithms, including Decision Tree (DT), Logistic Regression (LR), Random Forest (RF), Gaussian Naive Bayes (NB), and Support Vector Machine (SVM), were employed for result comparison and selection of the most accurate algorithm. RESULTS All algorithms demonstrated good performance, but the Random Forest (RF) algorithm stood out, achieving 100% accuracy in predicting the success/unsuccess status of the device. While the results of this research are specific to the collected data from EKG devices, the developed algorithms can be applied to other similar datasets, offering opportunities for broader use in the medical environment. CONCLUSION Implementing machine learning algorithms for automated systems in healthcare institutions can significantly enhance the quality of patient diagnosis and treatment. Additionally, these systems can optimize costs associated with managing medical devices. Improved post-market surveillance using ML can address challenges related to ensuring device reliability and safety.

BACKGROUND Atherosclerosis of the coronary arteries is a chronic, progressive condition characterized by the buildup of plaque within the arterial walls. Coronary artery disease (CAD), more specifically coronary atherosclerosis (CATS), is one of the leading causes of death worldwide. Computational modeling frameworks have been used for simulation of atherosclerotic plaque progression and with the advancement of agent-based modeling (ABM) the simulation results became more accurate. However, there is a need for optimization of resources for predictive modeling, hence surrogate models are being built to substitute lengthy computational models without compromising the results. OBJECTIVE This study explores the development of a surrogate model for atherosclerotic plaque progression using ABM simulation data. METHOD The dataset used for this study contains samples from latin-hypercube sampling based generated simulation parameters used in conjunction with 15 patient-specific geometries and corresponding plaque progression data. The developed surrogate model is based on deep learning using artificial neural networks (ANN). RESULTS The surrogate model achieved an accuracy of 95.4% in benchmarking with the ABM model it was built upon which indicates the robustness of the framework. CONCLUSION Adoption of surrogate models with high accuracy in practice opens an avenue for utilization of high-fidelity decision support systems for predicting atherosclerotic plaque progression in real-time.

With the advancement of Artificial Intelligence (AI), clinical engineering has witnessed transformative opportunities, enabling predictive maintenance of medical devices, optimization of healthcare workflows, and personalized patient care. Respiratory equipment plays a vital role in modern healthcare, supporting patients with compromised or impaired respiratory capacities. However, ensuring the reliability and safety of these devices is crucial to prevent adverse events and ensure patient well-being. This study aims to explore machine learning techniques to enhance predictive maintenance for mechanical ventilators. The dataset used for this study contains information about 1350 entries of mechanical ventilators, made by 15 different manufacturers and available in 30 distinct models. Different machine learning algorithms, including Logistic Regression, Decision Trees, Random Forest, K-nearest Neighbors, Support Vector Machines, Naive Bayes, and XG Boost are developed and tested in terms of their performance in predicting mechanical ventilator failures. The ensemble methods, particularly Random Forest and XGBoost, have proven to be more adept at handling the complexities of the dataset. The Decision Tree and Random Forest models both showed remarkable accuracies of approximately 0.993, while K-Nearest Neighbors (KNN) performed exceptionally with near perfect accuracy. Adoption of automated systems based on artificial intelligence will help in overcoming challenges of ensuring quality of MDs that are already being used in healthcare institutions. Implementing machine learning-based predictive maintenance can significantly enhance the reliability of mechanical ventilators in healthcare settings.

Accurate ear counting is essential for determining wheat yield, but traditional manual methods are labour-intensive and time-consuming. This study introduces an innovative approach by developing an automatic ear-counting system that leverages machine learning techniques applied to high-resolution images captured by unmanned aerial vehicles (UAVs). Drone-based images were captured during the late growth stage of wheat across 15 fields in Bosnia and Herzegovina. The images, processed to a resolution of 1024 × 1024 pixels, were manually annotated with regions of interest (ROIs) containing wheat ears. A dataset consisting of 556 high-resolution images was compiled, and advanced models including Faster R-CNN, YOLOv8, and RT-DETR were utilised for ear detection. The study found that although lower-quality images had a minor effect on detection accuracy, they did not significantly hinder the overall performance of the models. This research demonstrates the potential of digital technologies, particularly machine learning and UAVs, in transforming traditional agricultural practices. The novel application of automated ear counting via machine learning provides a scalable, efficient solution for yield prediction, enhancing sustainability and competitiveness in agriculture.

Analysis of data from incident registries such as MAUDE has identified the need to improve surveillance and maintenance strategies for infusion pumps to enhance patient and healthcare staff safety. The ultimate goal is to enhance infusion pump management strategies in healthcare facilities, thus transforming the current reactive approach to infusion pump management into a proactive and predictive one. This study utilized real data collected from 2015 to 2021 through the inspection of infusion pumps in Bosnia and Herzegovina. Inspections were conducted by the national laboratory in accordance with the Legal Metrology Framework, accredited to ISO 17020 standard. Out of 988 samples, 790 were used for model training, while 198 samples were set aside for validation (20% of the dataset). Various machine learning algorithms for binary classification of samples (pass/fail status) were considered, including Logistic Regression, Decision Tree, Random Forest, Naive Bayes, and Support Vector Machine. These algorithms were chosen for their ability to handle large datasets and potential for high prediction accuracy. Through detailed analysis of the achieved results, it was found that all applied machine learning methods yielded satisfactory results, with accuracy ranging from 0.98% to 1.0%, precision from 0.99% to 1%, sensitivity from 0.98% to 1.0%, and specificity from 0.87% to 1.0%. However, Decision Tree and Random Forest methods proved to be the best, both due to their maximum achieved values of accuracy, precision, sensitivity, and specificity, and due to result interpretability. It has been established that machine learning methods are capable of identifying potential issues before they become critical, thus playing a crucial role in predicting the performance of infusion pumps, potentially enhancing the safety, reliability, and efficiency of healthcare delivery. Further research is needed to explore the potential application of machine learning algorithms in various healthcare domains and to address practical issues related to the implementation of these algorithms in real clinical settings.

Healthcare institutions throughout the world rely on medical devices to provide their services reliably and effectively. However, medical devices can, and do sometimes fail. These failures pose significant risk to patients. One way to address these issues is through the use of artificial intelligence for the detection of medical device failure. This goal of this study was to develop automated systems utilising machine learning algorithms to predict patient monitor performance and potential failures based on data collected during regular safety and performance inspections. The system developed in this study utilised machine learning techniques as its core. Throughout the study four algorithms were utilised. These algorithms include Decision Tree, Random Forest, Linear Regression and Support Vector Machines. Final results showed that Random Forest algorithms had the best performance on various metrics among the four developed models. It achieved accuracy of 94% and precision and recall of 70% and 93% respectively. This study shows that use of systems like the one developed in this study have the potential to improve management and maintenance of medical devices.

BACKGROUND Poorly regulated and insufficiently maintained medical devices (MDs) carry high risk on safety and performance parameters impacting the clinical effectiveness and efficiency of patient diagnosis and treatment. After the MD directive (MDD) had been in force for 25 years, in 2017 the new MD Regulation (MDR) was introduced. One of the more stringent requirement is a need for better control of MD safety and performance post-market surveillance mechanisms. OBJECTIVE To address this, we have developed an automated system for management of MDs, based on their safety and performance measurement parameters, that use machine learning algorithm as a core of its functioning. METHODS In total, 1997 samples were collected during the inspection process of defibrillator inspections performed by an ISO 17020 accredited laboratory at various healthcare institutions in Bosnia and Herzegovina. This paper presents solution developed for defibrillators, but proposed system is scalable to any other type of MDs, both diagnostic and therapeutic. RESULTS Various machine learning algorithms were considered, including Decision Tree (DT), Random Forest (RF), Naïve Bayes (NB) and Logistic Regression (LR). In addition, random forest regressor and XG Boost algorithms were tested for their predictive capabilities in the field of defibrillator output error prediction. These algorithms were selected because of their ability to handle large datasets and their potential for achieving high prediction accuracy. The highest accuracy achieved on this dataset was 94.8% using the Naive Bayes algorithm. The XGBoost Regressor with its r2 of 0.99 emerged as a powerful tool, showcasing exceptional predictive accuracy and the ability to capture a substantial portion of the dataset's variability. CONCLUSION The results of this study demonstrate that clinical engineering (CE) and health technology management (HTM) departments in healthcare institutions can benefit from proposed automatization of defibrillator maintenance scheduling in terms of increased safety and treatment of patients, on one side, and cost optimization in MD management departments, on the other side.

Neurological impairment disorders in fetuses, such as cerebral palsy, epilepsy, and autism spectrum disorder, can arise from numerous factors impacting the development of the fetal nervous system. Although diagnosing these disorders early is difficult, it is essential for prompt intervention. Recent progress in deep learning and ultrasound technology offers the potential to create a tool for early detection. Development of the TRUEAID system is based on combining the meticulously tuned Kurjak Antenatal Neurodevelopmental Test (KANET) with a sophisticated convolutional neural network for construction of an AI empowered ultrasound module capable of automated diagnostic decision support in the field of fetal neurodevelopmental risk assessment. The model's performance was evaluated using accuracy metrics, precision, sensitivity, specificity, F1 score, and Mathesson Correlation Coefficient (MCC). The custom CNN architecture achieved an overall accuracy of 93.83%. This pilot study lays the foundation for AI-based fetal neurobehavioral assessment, providing a promising tool for the early detection of fetal neurological impairment disorders. The research holds implications for improving outcomes for affected children and making advanced diagnostic capabilities accessible in diverse healthcare settings.