In recent advancements in robotics, Artificial Intelligence (AI) methods such as Deep Learning, Deep Reinforcement Learning (DRL), Transformers, and Large Language Models (LLMs) have significantly enhanced robotic capabilities. Key AI models driving advancements in robotic vision include Convolutional Neural Networks (CNNs), Vision Transformers (ViTs), the DEtection Transformers (DETR), the YOLO family of algorithms, segmentation techniques, and 3D vision technologies. Deep Reinforcement Learning (DRL), an AI technique where agents learn optimal behaviors through trial and error interactions with their environment, enables robots to perform complex tasks autonomously. Transformers, originally developed for natural language processing, have been adapted to robotics for tasks involving sequence prediction and data understanding, improving perception and decision-making processes. LLMs leverage vast amounts of text data to enhance robot-human interaction, enabling robots to understand and generate human-like language, thus improving their communicative and collaborative abilities in various applications. The integration of these AI methods enhances the adaptability, efficiency, and overall performance of robotic systems, paving the way for more sophisticated and intelligent autonomous agents.

Objective The aim was to test the Belgrade age formula based on the calculation of open apices of two permanent mandibular teeth on a Bosnian children population and compare its accuracy with European formula. Material and methods We included 412 panoramic images of children (204 female and 208 male) 7 to 13 years of age. We assessed the performance of both methods (the European formula and the BAF) and compared their results in both sexes. Results The results showed a high point of average understanding between the age estimated by chronological age and the European formula (ICC=0.927, 95% CI 0.904–0.944, p<0.001)., BAF also confirmed a high point of agreement with chronological age in boys (ICC=0.941, 95% CI 0.922–0.955, p<0.001) and girls (ICC=0.913, 95% CI 0.886–0.934, p<0.001). BAF was better than the European formula in estimating age in males (0.4448±0.9135 vs. 0.9807±0.9422). Conclusion The Belgrade Age Formula (BAF) demonstrates comparable accuracy to the European formula for age determination in Bosnian children, while offering the advantage of being easier and faster to use. This makes the BAF a practical alternative in clinical and research settings where efficiency and reliability are essential.

Although deep learning (DL) algorithms have been proved to be effective in diverse research domains, their application in developing models for tabular data remains limited. Models trained on tabular data demonstrate higher efficacy using traditional machine learning models than DL models, which are largely attributed to the size and structure of tabular datasets and the specific application contexts in which they are utilized. Thus, the primary objective of this paper is to propose a method to use the supremacy of Stacked Bidirectional LSTM (Long Short-Term Memory) deep learning algorithms in pattern discovery incorporating tabular data with customized 3D tensor modeling in feeding neural networks. Our findings are empirically validated using six diverse, publicly available datasets each varying in size and learning objectives. This paper proves that the proposed model based on time-sequence DL algorithms, which were generally described as inadequate when dealing with tabular data, yields satisfactory results and competes effectively with other algorithms specifically designed for tabular data. An additional benefit of this approach is its ability to preserve simplicity while ensuring fast model training also with large datasets. Even with extremely small datasets, models can be applied to achieve exceptional predictive results and fully utilize their capacity.

The deployment of diverse data-generating technologies in livestock farming holds the promise of early disease detection and improved animal well-being. In this paper, we combine routinely collected dairy farm and herd data with weather and high frequency sensor data from 6 farms to predict new lameness events in various future periods, spanning from the following day to 3 weeks. A Random Forest classifier, using input features selected by the Boruta Algorithm, was used for the prediction task; effects of individual features were further assessed using partial dependence plots. We achieve precision scores of up to 93% when predicting lameness for the next 3 weeks and when using information from the last 3 weeks, combined with a balanced accuracy of 79%. Removing sensor data results have tendency to reduce the precision for predictions, especially when using information from the last one,2 or 3 weeks. Moving to a larger data set (without sensor data) of 44 farms keeps the similar balanced accuracy but reduces precision by more than 30%, revealing a substantial a trade-off in model quality between false positives (false lameness alerts) and false negatives (missed lameness events). Sensor data holds promise to further improve the precision of these models, but can be partially compensated by high resolution data from other systems, such as automated milking systems.

Abstract This paper introduces a novel method that leverages artificial neural networks to estimate magnetic flux density in the proximity of overhead transmission lines. The proposed method utilizes an artificial neural network to estimate the parameters of a mathematical model that describes the magnetic flux density distribution along the lateral profile for various configurations of overhead transmission lines. The training target data is acquired using the particle swarm optimization algorithm. A performance comparison between the proposed method and the Biot-Savart law-based method is conducted using an extensive test dataset. The resulting coefficient of determination and mean square error values demonstrate the successful application of the proposed method for a range of different spatial arrangements of phase conductors. Furthermore, the performance of the proposed method is thoroughly assessed on multiple test cases. The practical relevance of the proposed method is highlighted by contrasting its results with the field measurements obtained in the proximity of a 400 kV overhead transmission line.

Objective The objective of this study was to evaluate the root canal morphology of third molars in the Bosnia-Herzegovina population. Materials and methods A total of 241 extracted third molars (105 maxillary and 136 mandibular) were subjected to a clearing procedure. The specimens were categorized into ten groups based on the Alavi classification for maxillary third molars (MaxTMs), and six groups were based on the Gulabivala classification for mandibular third molars (ManTMs). Root canal type according to the Vertucci classification, the presence and position of lateral canals, and intercanal communication were analyzed using a stereomicroscope x15. Results MaxTMs had three roots in 77.13% of the samples. Among MaxTMs, the most common morphology was three fused roots (33.33%) and Vertucci’s type VIII (54. 28% of samples in Alavi’s Group IV). 60.29% of ManTMs have two separate roots (Gulabivala's Groups II and III). The most prevalent types in mesial roots were type I (41.46% in Group II) and type IV (48.78% in Group III), although type I predominated in distal roots (91.24% and 100% in Groups II and III, respectively). Conclusion Single-rooted third molars usually have a root canal morphology that is more favorable for endodontic treatment. In contrast, third molars with fused roots often have more complex root canal morphology.

In the last ten years, the development and research of advanced technologies, as well as their application in all segments of society, have led to major changes and reshaping of the new world. New innovations are occurring on a daily basis, but their application is not going fast enough due to the rigid infrastructure. However, in order to secure an optimal future, we all have to adapt to the changes that are coming. The developed countries have adopted the strict implementation of advanced technologies of Industry 4.0, some of which include: Internet of Things (IoT), Big Data, Cloud Computing, smart sensors, Radio Frequency Identification (RFID), 3D printing, advanced security systems, Virtual and Augmented Reality (VAR), etc. Robotics is the basic and first technology that has been implemented since the 60s of the last century, with artificial intelligence coming in the spotlight in the last ten years. Artificial intelligence is becoming a key to the development of advanced robots, as it enables them to adapt to unpredictable situations, to learn from experience and make intelligent decisions.Robots use AI to process sensor data, navigate, recognize objects, plan paths and interact with the environment. In short, artificial intelligence enables robots to be smart, whereas robotics uses AI to create autonomous and useful devices. This symbiosis contributes to progress in many industries, including healthcare, manufacturing and transportation. Artificial intelligence (AI) and robotics are two key fields that complement each other. The paper presents the trend of applied and approved patents in artificial intelligence and robotics, as well as an example of the use of artificial intelligence in advanced robots to perform certain tasks. Artificial intelligence (AI) is having an increasing impact on robotics, opening up many possibilities.

Using AI through industries and business processes is increasingly becoming the subject of theorists and practitioners. In the HRM process, the use of AI gives companies numerous advantages in employee performance, and processes, but also presents them with organizational, financial, technical, legal, and personnel challenges. This paper explores the application of AI systems in recruitment and selection through gamification strategies, people analytics, talent intelligence, AI platforms, video interviews, and conversational AI. It provides an overview of the benefits and challenges associated with their implementation. Additionally, the paper delves into ethical considerations and legislation, focusing on the EU Act, domestic laws, and ISO AI standards. The primary goal of this paper is to provide a comprehensive understanding of AI's role in HR processes and the complexities of implementing AI solutions in recruitment and selection.

U radu je predstavljen postupak izbora najpogodnijeg numeričkog modela za utvrđivanje indeksa staništa (SI – site index) kao apsolutne mjere proizvodnog potencijala (boniteta) staništa jednodobnih nenjegovanih sastojina bijelog bora na karbonatnim supstratima u BiH. Objekat istraživanja su predstavljale jednodobne nenjegovane sastojine bijelog bora različitih taksacionih i stanišnih karakteristika. Metodom privremenih oglednih parcela prikupljeno je više općih i taksacionih podataka, a zatim su njihovom obradom i analizom utvrđeni najvažniji strukturni i proizvodni parametri sastojina odvojeno po relativnim visinskim bonitetnim klasama staništa (RB). Za utvrđivanje numeričkog modela za procjenu indeksa staništa (SI) primijenjene su metode korelacione i regresione analize, a za predstavljanje veličina osnovnih taksacionih elemenata prema veličinama SI grafička metoda. U cilju predstavljanja veličina osnovnih taksacionih elemenata po utvrđenim SI klasama uspostavljena je korelaciona veza između SI50 (pri starosti od 50 godina)i postojećih relativnih bonitetnih klasa (RB) jednodobnih sastojina bijelog bora. Ova veza je poslužila za izradu proizvodne diferencijacije staništa jednodobnih sastojina bijelog bora na karbonatnim susptratima u BiH koja omogućava prikaz veličina osnovnih taksacionih elemenata ovih sastojina zavisno od starosti i SI50. Poređenjem utvrđenih rezultata istraživanja s odgovarajućim rezultatima drugih autora zaključeno je da su jednodobne sastojine bijelog bora u BiH srednje produktivne.

This study aimed to explore how Dab1 functional silencing influences the expression patterns of different connexins in the developing yotari (yot) mice eyes as potential determinants of retinogenesis. Using immunofluorescence staining, the protein expression of Dab1, Reelin, and connexin 37, 40, 43, and 45 (Cx37, Cx40, Cx43, and Cx45) in the wild-type (wt) and yot eyes at embryonic days 13.5 and 15.5 (E13.5 and E15.5) were analyzed. Different expression patterns of Cx37 were seen between the wt and yot groups. The highest fluorescence intensity of Cx37 was observed in the yot animals at E15.5. Cx40 had higher expression at the E13.5 when differentiation of retinal layers was still beginning, whereas it decreased at the E15.5 when differentiation was at the advanced stage. Higher expression of Cx43 was found in the yot group at both time points. Cx45 was predominantly expressed at E13.5 in both groups. Our results reveal the altered expression of connexins during retinogenesis in yot mice and their potential involvement in retinal pathology, where they might serve as prospective therapeutic targets.

In hip fracture patients, who are mostly elderly, preexisting anemia can be worsened when combined with trauma and surgery. To this date, there is no unequivocal approach about transfusion thresholds. We analyzed hemoglobin (Hb) and hematocrit (Hct) levels at three time points in surgical patients with proximal femoral fractures (PFF) to see which levels were triggers for transfusions and whether transfusions were related to mortality after hospital discharge. A total of 956 patients were operated on from 1 January 2021 to 31 December 2022 at the University Hospital of Split and included in the study. There were more women (74%); 47% patients had admission Hb < 120 g/L. Transfusion was given preoperatively to 88, intraoperatively to 74 and postoperatively to 309 patients. Transfusion thresholds were as follows: Hb 84 g/L preoperatively, 99 intraoperatively and 83 postoperatively. After hospital discharge, 10.79% of patients died within the 1st month and 23% within 6 months. In the group of non-survivors, 60% of patients had admission Hb ≤ 117 g/L and the proportion of patients transfused preoperatively was two times higher. Preoperative transfusion thresholds could be set to higher levels for patients with surgically treated PFF. However, that could increase mortality even more. Further investigation is necessary.

Abstract Introduction The aim of this research was to examine the existence of broncho-obstruction of the airways in patients suffering from chronic obstructive pulmonary disease, as well as changes in resistance values in this part of the bronchial tree after inhalation of salbutamol. Methods This study was designed as a prospective, interventional clinical trial that included a sample of 147 patients suffering from COPD. Patients were stratified into four groups of thirty patients each based on the severity of airflow limitation (based on the post-bronchodilator FEV1 value), according to the GOLD grade. The test was conducted at the University Clinical Center Sarajevo, Clinic for Lung Diseases and Tuberculosis “Podhrastovi” Results The results of our research showed that the average values of FEV1 compared to the predicted values of this parameter in subjects in the GOLD 4 group before the administration of salbutamol were statistically significantly lower than the average values of FEV1 in other subjects of the GOLD group. After the administration of salbutamol, a statistically significant increase in the value of FEV1 was registered in all tested groups. When the response to salbutamol was compared among the GOLD groups, it was assessed that the difference in the percentage increase in predicted FEV1 values after the administration of salbutamol among the tested groups was not statistically significant. Conclusion Bronchodilation with salbutamol and additional ipratropium had a significant effect on both mentioned parameters, especially in the COPD group, which speak in favor of the presence of increased peripheral resistance in all groups of patients. A statistically significant bronchodilator response was obtained in GOLD 1 and GOLD 2 groups, i.e. in groups of patients with milder forms of the disease.

: Bijeli bor (Pinus sylvestris L.) jedna je od najvažnijih i najistraženijih vrsta drveća u evropskim šumama. Cilj istraživanja je utvrditi mogućnosti povećanja proizvodnih mogućnosti i kvaliteta bijelog bora kroz ranu selekciju u testu provenijencija u Bosni i Hercegovini, a rezultati će se koristiti u svrhu što boljeg korištenja staništa pogodnih za bijeli bor. Za potrebe ovog istraživanja 2021. godine mjerene su visine i prečnici u prsnoj visini bijelog bora u testu provenijencije u Žepču, te brojane grane u pršljenu. Test je uspostavljen 2012. godine sadnjom dvogodišnjih sadnica iz 14 evropskih provenijencija (po tri iz Austrije i Italije i po jedna iz Bosne i Hercegovine, Njemačke, Poljske, Rumunije, Slovačke, Norveške, Škotske i Ukrajine). Izračunate su i zapremine, a zatim provedena analiza varijanse i deskriptivna analiza podataka po provenijencijama. Zatim je selekcionirano pet provenijencija sa najvećim prosječnim vrijednostima istraživanih svojstava i izračunate su razlike između prosječnih vrijednosti za selekcionirane provenijencije i ukupnog prosjeka, kao i razlike između prosječnih vrijednosti za provenijenciju sa najvećim prosječnim vrijednostima i ukupnog prosjeka za sva istraživana svojstva. Podaci su obrađeni pomoću Excel 2013 i SPSS 20.0. Analiza varijanse pokazala je statistički značajne razlike između provenijencija po svim istraživanim svojstvima. Najveći prosječan prečnik stabala starih 11 godina imala je provenijencija iz Ukrajine, a najveću prosječnu visinu i zapreminu jedna od provenijencija iz Austrije. Razlika između prosjeka pet provenijencija sa najvećim prosječnim vrijednostima i ukupnog prosjeka po svojstvu zapremine iznosila je 17,4% od ukupnog prosjeka, a razlika između provenijencije sa najvećom prosječnom zapreminom i ukupnog prosjeka iznosila je 41% od ukupnog prosjeka. Rezultati će biti korišteni u procesima selekcije provenijencija bijelog bora sa dobrim visinskim i debljinskim prirastom.

Crna joha (Alnus glutinosa [L.] Gaertn.) listopadno je drvo iz porodice brezovki (Betulaceae) koje se uglavnom javlja u nizijskim područjima u blizini vodnih tijela različitog režima plavljenja, često prateći riječne i manje vodene tokove, formirajući zajednice azonalnog karaktera u Bosni i Hercegovini. Cilj istraživanja je razvijanje predikcijskog modela rasprostranjenosti staništa crne johe u Bosni i Hercegovini na osnovu poznatih i utvrđenih lokacija staništa crne johe i podataka WorldClim baze. Kao osnovni materijal istraživanja korišteni su podaci sa vlastitih terenskih snimanja i iz drugih relevantnih dostupnih izvora. Klimatski podaci su preuzeti sa WorldClim baze podataka u obliku rasterskih slojeva za 19 bioklimatskih varijabli. Modeliranje je provedeno koristeći metod maksimalne entropije integrisan u MaxEnt računarski program. Izbor modela se zasnivao na predikcijskoj tačnosti modela i koherentnosti sa distribucijom staništa crne johe trenutnog i prethodnih istraživanja. Dobiveni rezultati su potvrdili model veoma dobre predikcijske tačnosti na temelju kalibracije i validacije podataka subseta (AUC > 0,85). Varijable BIO6 (minimalna temperatura najhladnijeg mjeseca) i BIO14 (oborine mjeseca s najmanje padavina) imaju statistički najveći značaj i uticaj za model prostorne rasprostranjenosti ekološke niše crne johe. S obzirom na važne funkcije zajednica crne johe u prevenciji erozije, zaštiti od poplava i očuvanju biološke raznolikosti, dobiveni rezultati mogu doprinijeti upravljanju staništima crne johe. Rezultati predikcije staništa mogli bi se koristiti za daljnja istraživanja vezana za klimatske promjene i praćenje stabilnosti ekosistema.