The aim of this study was to explore diplopia as a symptom of undetected COVID-19 infection or as a possible side effect of COVID-19 vaccination. We examined 380 patients with diplopia admitted to the Department of Ophthalmology of the University Hospital Centre Sestre milosrdnice in Zagreb, Croatia, from July 2020 to June 2022. After excluding patients with confirmed organic underlying diplopia causes or monocular diplopia, we linked the patient information with the national COVID-19 and vaccination registries. Among the 91 patients included in this study, previously undetected COVID-19 infection as the possible cause of diplopia was confirmed in five of them (5.5%). An additional nine patients (9.9%) were vaccinated within one month from the onset of their symptoms, while the remaining 77 had neither and were therefore considered as controls. The breakdown according to the mechanism of diplopia showed no substantial difference between the vaccinated patients and the controls. We detected marginally insignificant excess abducens nerve affection in the COVID-positive group compared with that in the controls (p = 0.051). Post-vaccination diplopia was equally common in patients who received vector-based or RNA-based vaccines (21.4 vs. 16.7%; p = 0.694). COVID-19 testing should be performed for all cases of otherwise unexplained diplopia. The risk of post-vaccination diplopia was similar in both types of vaccines administered, suggesting a lack of evidence linking specific vaccine types to diplopia.

abstract:Population genetic studies have shown that the Bosnian-Herzegovinian (B&H) population is a part of the European gene pool, but there has been limited information on the genetic structure of ancient B&H populations. This study aimed to determine the frequency and distribution of mitochondrial DNA (mtDNA) haplogroups for a medieval Bosnian population. Thirty-four samples, excavated from medieval necropolises located within the borders of medieval Bosnia, were analyzed. Sequencing of the mtDNA hypervariable segment 1 (HVS1) region and RFLP analysis were performed for haplogroup determination. All 32 samples were identified as haplogroup H, with subhaplogroups H2a and H5 in 30 and 2 samples, respectively. The frequency of the H haplogroup was significantly different between the studied samples and previous studies of contemporary B&H populations, where the H haplogroup frequency was approximately half that of the ancient population studied here. A significant difference in H haplogroup frequency compared with other medieval populations outside of Bosnia was also observed: the ancient B&H population is most similar to ancient Italians. These results provide insight into the mitochondrial landscape of populations that inhabited the territory of present-day Bosnia and Herzegovina in the Middle Ages. Our study reveals that inhabitants of medieval Bosnia carried genetic lineages that exist today in B&H populations, suggesting continuity of mtDNA haplogroups over a long period of time, regardless of various historical demographic events that shaped the genetic structure of the modern B&H population.

Concrete structures, such as bridges or viaducts, play an important role in global road infrastructure. These types of structures are relatively expensive to build and they are susceptible to outer external influences, which in time deteriorate and lead to the reduction of their structural resistance. To reduce this effect, regular inspection is needed, which is often done manually by using specialized equipment to reach certain parts of bridges and viaducts. This process is both expensive and dangerous for the inspectors to conduct. Within the research project ASAP (Autonomous System for Assessment and Prediction of Infrastructure Integrity) in order to overcome these challenges, we have developed a prototype of a wall-climbing robot (WCR) for nondestructive testing (NDT). In this paper, different iterations of the developed WCR prototypes are presented. In four consecutive prototype designs, we have evaluated and upgraded the adhesion and locomotion system. Finally, a fifth prototype that carries the NDT equipment is presented. The final version of the WCR is equipped with robust and flexible adhesion that enables the robot to adhere to different types of surfaces. We have also addressed the challenges of integrating NDT equipment into the robot. To successfully conduct an inspection, besides the WCR, a safety system, control, and power systems are needed, which are further presented and discussed.

In this work we introduce the concept and method of so-called cooperative solar generation forecasting, where geographically close data sources are utilized in order to improve forecasting accuracy. We devised and examined various largescale one-hour-ahead artificial neural networks based solar generation forecasting scenarios to prove the benefits of cooperation. The introduced cooperative solar generation forecasting method showed significant improvement in forecasting accuracy, especially when combined with previous generation data, where a root mean square error reduction of at least 50% could be achieved in the majority of cases. We believe these results point to a scientific and economical benefit of international cooperation in solar generation forecasting.

Malware traffic classification (MTC) is a very important component of cyber security, and a number of the MTC techniques are based on deep learning (DL) with a strong capability of feature mining and classification. However, these DL-based MTC methods are heavily dependent on a large amount of network traffic samples. In the few-shot scenarios, these methods usually overfit and have poor classification performance. Considering that the update cycle of malware is faster and faster, and there are more and more types of malware, collecting enough training samples for all malware is very challenging, if not impossible. In this paper, a novel few-shot MTC(FS-MTC) method is proposed based on convolutional neural network (CNN) and model-agnostic meta-learning (MAML) algorithm. Specifically, the CNN is trained on samples from normal softwares by MAML rather than the conventional optimization methods, then the CNN is finetuned by a few samples from malware for MTC. Simulation results show that our proposed MAML-based FS-MTC can outperform the traditional MTC methods. The performance of our proposed method can reach up to 95.69%.

To address the problem of spectrum resources and transmitting power for vehicular networks, this paper proposes a resource allocation (RA) method based on dueling double deep-Q network (D3QN) reinforcement learning (RL). Due to the high mobility of the vehicle, the channel changes rapidly which makes it difficult to accurately collect high-accuracy channel state information at the base station and to perform centralized management. In response of this difficulty, we construct a multi-intelligence model, using Manhattan Grid Layout City Model as the basis of environment and with each vehicle-to-vehicle (V2V) link as an intelligence. They work together to interact with the environment, receive appropriate observations, get rewards, and finally learn to improve the allocation of power and spectrum to enable users to achieve a better entertainment experience and a safer driving environment. Experimental results demonstrate that with proper training mechanism and reward function construction, cooperation among multiple intelligence can be performed in a distributed manner, with improvements in both the capacity of total vehicle-to-infrastructure links and the effective payload delivery success rate of the V2V links compared to common Q-network.

Specific emitter identification (SEI) is developed as a potential technology against attackers in cognitive radio networks and authenticate devices in Internet of Things (IoT). It refers to a process to discriminate individual emitters from each other by analyzing extracted characteristics from given radio signals. Due to the strong capability of deep learning (DL) in extracting the hidden features of data and making classification decision, deep neural networks (DNNs) have been widely used in the SEI. Considering the insufficiently labeled training dataset and large unlabeled training dataset, we propose a novel SEI method using semi-supervised (SS) learning framework, i.e., metric-adversarial training (MAT). Specifically, two object functions (i.e., cross-entropy (CE) loss combined with deep metric learning (DML) and CE loss combined with virtual adversarial training (VAT)) and an alternating optimization way are designed to extract discriminative and generalized semantic features of radio signals. The proposed MAT-based SS-SEI method is evaluated on an open source large-scale real-world automatic-dependent surveillance-broadcast (ADS-B) dataset. The simulation results show that the proposed method achieves a better identification performance than four latest SS-SEI methods.

Radio frequency fingerprint (RFF) is regarded as a key technology in physical layer security in various wireless communications systems. Deep learning (DL) has achieved great success in the field of signal identification, particularly in improving performance and eliminating manual feature extraction. However, the training cost of these DL-based methods is usually large. It is unwise to retrain the network with whole data when it comes to new data. Therefore, we propose a novel RFF identification method based on incremental learning (IL), which uses continuous data stream to update the identification model, constantly. Experimental results show that with the increase of increment times, the accuracy of the proposed IL-based method gradually approaches the performance of joint training, and finally reaches 96.79%, which is only 1.9% lower than the performance upper bound.

The fast development of intelligent wireless communications enables many devices to access various networks. It often leads to the security risks of malicious access of illegal devices. To ensure a secure and reliable wireless access, it is necessary to identify illegal devices and prevent their attacks accurately. To improve the performance of specific emitter identification (SEI), this paper proposes a multi-scale convolution neural network (MSCNN) based on convolution layers of three branches with different convolution kernel sizes. MSCNN extracts radio frequency fingerprints (RFF) in three receptive fields through different convolution kernels. We verify the identification accuracy using the RF signals conforming to long term evolution (LTE) standard. The experimental results show that our proposed MSCNN-based SEI method can improve the absolute accuracy by 15% and the relative accuracy by 22% in perfect communication environment. In addition, we verify the robustness of proposed MSCNN by comparing identification performance in imperfect environment. Simulation results show that the proposed MSCNN can extract more hidden features through convolution kernels of different sizes, and thus achieves better SEI performance than existing methods.

Automatic modulation recognition (AMR) technique plays an important role in the identification of modulation types of unknown signal of integrated sensing and communication (ISAC) systems. Deep neural network (DNN) based AMR is considered as a promising method. Considering the complexity of a typical ISAC system, devising the DNN manually with limited knowledge of its various classifications will be very tasking. This paper proposes a neural architecture search (NAS) based AMR method to automatically adjust the structure and parameters of DNN and find the optimal structure under the combination of training and constraints. The proposed NAS-AMR method will improve the flexibility of model search and overcome the difficulty of gradient propagation caused by the non-differentiable quantization function in the process of back propagation. Simulation results are provided to confirm that the proposed NAS-AMR method can identify the modulation types in various ISAC electromagnetic environments. Furthermore, compared with other fixed structure networks, our proposed method delivers the highest recognition accuracy, under the condition of low parameters and floating-point operations (FLOPs).

In the past decade, we have witnessed the emergence of a large number of different computer-based animations and simulations that have the goal to foster better learning of different physics topics. Past studies have shown many benefits of animations and simulations, but for their efficient usage it is very important that teachers are well educated in the teaching material and usage of selected visualizations. Furthermore, studies have proven that augmented reality technology has a potential to reduce cognitive load and improve the quality of physics lectures. Many of these visualizations are generally designed for targeted physics phenomena, and sometimes it is not easy to address specific students’ misconceptions. In this paper, we will present augmented reality animations and a simulation that can generally be useful for teaching about counterintuitive aspects of rolling motion, and specifically address students’ misconceptions about rolling friction and velocity in contact with the ground.

Background: History of pharmaceutical industry in Bosnia and Herzegovina (B&H) has its roots from 1951. Importance of domestic industry not just from economical aspect but also from public health perspective and as scientific base has not been evaluated previously. Objective: The aim of this article was to provide the review of the pharmaceutical industry developments in Bosnia and Herzegovina, its roots, current position and future perspectives.. Methods: Research of published scientific papers as well other documents and archives of pharmaceutical manufacturers has been conducted. We have also analysed market trends focusing on domestic producers. Results and Discussion: Over more than seventy years of B&H pharmaceutical industry has been developing. During Yugoslavia only two companies existed of which one, Bosnalijek is still present, while Sanofarm has been closed. After 1996, expansion of domestic manufacturers started and today six companies are present. They are mainly oriented to generic drugs production in different forms. Total market share of domestic producers in B&H is 20-25% which is relatively low comparing to other countries. Many of domestic manufacturers are exporting their products to some of the most demanding markets in Europe and Middle East. Conclusion: Long history of domestic drug manufacturers in B&H gives solid legacy for future developments. Importance of local producers has been confirmed during war in B&H and COVID-19 pandemic as a crisis situation, mainly from public health perspective and sustainable supply of essential medicines. Higher support by state and collaboration with academia in order to expand portfolio, especially in area of biologic medicines is required in future.

<jats:p />