Background: The basic model of a neural network is the physical structure of the human brain. The idea is simple: if a sufficiently powerful artificial network of synapses is created, the results of functioning similar to the human brain can be expected . Neural network or neurocomputer is one form of artificial intelligence system implementation, which represents a system consisting of a certain number of interconnected processors or nodes, or process elements that we call artificial neurons. Objective: The aim of this article was to draw attention to the necessity of taking about a detailed history of neural networks which are based on copying information processing in synaptic networks in neurons and synapses in the brain. Methods: The author searched scientific literature: the books and published papers deposited in index databases as source for description and explanation the architecture of a neural networks and its represents a specific connection of neurons into a single unit. Resuls and Discussion: The structure of a neural network differs in the number of layers. The first layer is called the input layer, and the last is the output layer, while the layers in between are called hidden layers. Most often, there are three of them. The first layer, or the input layer, is the only layer that receives data from the external environment, the next (hidden) layer forwards the relevant data to the third (output) layer. At the output of the third layer, we get the final result. More complex neural networks have more hidden layers. The layers are fully interconnected. Layers communicate by connecting the output of each neuron from the previous layer to the inputs of all neurons in the next layer. So, each node has several inputs and one output. The strength of the connections by which the neurons are connected is called the weight factor. Conclusion: In the past several books were written and several companies dealing with neurocomputers were founded. The computers were able to successfully adjust the weight coefficients, did not achieve significant practical results. Although neural networks have had an unusual history, they are still in an stage of development. Today, neural networks find a very wide range of applications in various practical areas.

Eryngium amethystinum L is a plant which grows in Balkan and Apennine peninsulas. The aerial parts, roots and fruits are used in ethnopharmacology of Italy and Western Balkan countries. Traditional preparations of E amethystinum were used in treatment of oedema, malaria and gastrointestinal diseases. Essential oils and extracts were isolated from aerial parts and roots of E amethystinum and examined by antimicrobial, antioxidative and cytotoxic assays. Gas-chromatography analysis showed predominance of germacrene D and spathulenol in essential oils of E amethystinum aerial parts, which have contributed to strong cytotoxic activity, while the methanolic extract exhibited strong antioxidative and antimicrobial activity. This article summarises all existing knowledge regarding E amethystinum, its chemical composition and pharmacological activity.

This paper presents a mathematical model for the calculation of transient current distribution in grounding systems, based on antenna theory. The modeling of grounding electrodes relies on a system of coupled integro-differential equations of the Pocklington type, applied to complex wire structures buried in a semi-space with finite conductivity. The Pocklington equation is derived directly from Maxwell’s equations, and the paper thoroughly describes the entire procedure, including the influence of the boundary between two media using Fresnel reflection coefficients. The system of equations is solved using the indirect boundary element method, resulting in the determination of current distribution along grounding structures of various geometries, which represents a fundamental parameter for analyzing the transient response of grounding systems.

Background: Suicide is a major public health issue and a leading cause of premature mortality worldwide. Assessing suicide risk remains challenging due to multifactorial risks. Objective. The aim of the study was to present ten-year trends in suicides in the Zenica-Doboj Canton and to describe the methods of suicide by gender, age, and year of occurrence, with the goal of better understanding this complex and deviant form of behavior. Methods: In the article has been included a retrospective analysis of suicide data collected from the suicide reporting form in the Zenica-Doboj Canton, obtained from the Institute for Health and Food Safety Zenica for the period 2015–2024. Results: Resuzlts are based on 147 processed cases of suicide over a ten-year period. The distribution of suicides by age group was: under 1 year: 0 (0%), 1–4: 0 (0%), 5–9: 0 (0%), 10–14: 0 (0%), 15–19: 4 (3%), 20–29: 16 (11%), 30–39: 13 (9%), 40–49: 20 (15%), 50–59: 26 (18%), 60–64: 27 (19%), 65–69: 9 (6%), 70–79: 20 (14%), 80+: 12 (8%). The average number of suicides per age group was x̄ = 11.3. Yearly suicide cases: 2015: 19 (12.9%), 2016: 19 (12.9%), 2017: 22 (14.9%), 2018: 29 (19.7%), 2019: 29 (19.7%), 2020: 8 (5.4%), 2021: 1 (0.68%), 2023: 0 (0%), 2024: 0 (0%). The average number of suicides per year was x̄ = 14.7. The gender distribution was 73% male and 27% female, with a male-to-female ratio of 2.7:1. The most common method was intentional self-harm by hanging, strangulation, and suffocation; 93 (63.2%). Conclusion. Understanding suicide trends, mechanisms, and methods in the Zenica-Doboj Canton can help in the development of early-prevention programs and prevention strategies, as well as in better understanding the contributing factors and suicidal individuals’ attitudes that lead to such fatal decisions.

This paper explores the impact of performance expectancy (PE) and effort expectancy (EE) on the behavioral intention (BI) of managers to implement environmental, social, and governance (ESG) practices in small and medium-sized enterprises (SMEs) in Bosnia and Herzegovina. Us ing the Technology Acceptance Model (TAM), the study analyzes how these factors influence decision-makers’ engagement in sustainable business activities. A survey of 247 managers was conducted to assess their BI, PE, and EE for each ESG component. The collected data were ana lyzed using structural equation modeling (SEM) to construct a second-order latent model that measures the combined effects of these factors. The study fills a gap in the literature by examin ing sector-specific differences in the factors influencing ESG implementation. The results reveal that both PE and EE significantly affect an organization’s intention to adopt ESG practices, with varying impacts between the manufacturing and service industries. The findings highlight the different challenges and complexities each industry faces in implementing ESG practices. While focusing on PE and EE, it is acknowledged that other factors such as organizational culture, stakeholder pressure, and industry-specific dynamics also play roles in ESG adoption. The main contribution of this work is the use of a second-order latent construct within the TAM model for ESG practices, offering a unique perspective on understanding behavioral intention.

Introduction Renal angiomyolipomas (AMLs) are neoplasms that can rarely rupture, causing hemorrhagic shock as the most serious complication. This pathological condition, (referring to AML) is classified as a benign tumor arising from the proliferation of epithelioid cells, consisting of fat tissue, blood vessels, and smooth muscle. Wunderlich syndrome describes a spontaneous, nontraumatic bleeding into the subcapsular or perirenal space. Most individuals with renal AML exhibit no symptoms and are often diagnosed incidentally, however, some may experience life-threatening complications such as rupture, hemorrhage, and circumstantial hypovolemic shock. Case outline Description of the clinical presentation of AML with rupture in a female patient with a brief overview of other cases of AML in the literature. Female patient, 68 years old, admitted for examination due to sudden severe pain in the abdomen with propagation to the right lumbar region accompanied by nausea and fatigue. After a complete physical examination, an abdominal ultrasound, and a CT scan, surgery was performed during which the right kidney was removed alongside the hematoma and the kidney envelopes, which were sent for pathohistological analysis. The result of the histopathological analysis confirmed that it was AML. Conclusion AMLs are benign neoplasms with potentially serious complications. The most serious complication of AML is rupture, leading to retroperitoneal hemorrhage, with tumor size being a significant risk factor. Considering the clinical importance of this potential complication, it is important to establish a swift and accurate radiological diagnosis, with the aim of timely therapeutic intervention and reduction of potential additional complications.

Timely and accurate defect detection is essential in the leather industry, as the quality of raw leather directly impacts both the usability and value of finished products. This paper provides a systematic overview of state-of-the-art solutions and proposes a novel approach for automated detection of leather surface defects using deep neural networks based on the Inception-V3 architecture. Five defect categories are introduced, focusing on their impact on leather quality. In addition, two deep neural network architectures were analyzed and implemented for defect detection and classification: a single-channel model and a multi-channel model with arbitration. The evaluation was carried out using a combination of a custom-developed dataset and publicly available datasets, assessed with standard performance metrics. Moreover, an image annotation tool was developed to facilitate precise defect labeling and the creation of variable-size datasets. Both models demonstrated promising results on the custom dataset, achieving accuracy rates exceeding 93%. The suggested methodology enhances the research domain of leather inspection automation by creating an openly accessible image dataset, performing a comparative analysis of detection models and creating software tools for data preparation. These contributions lay the foundation for further research in leather defect detection and potential industrial implementation.