In recent years, deep learning and especially deep neural networks (DNN) have obtained amazing performance on a variety of problems, in particular in classification or pattern recognition. Among many kinds of DNNs, the convolutional neural networks (CNN) are most commonly used. However, due to their complexity, there are many problems related but not limited to optimizing network parameters, avoiding overfitting and ensuring good generalization abilities. Therefore, a number of methods have been proposed by the researchers to deal with these problems. In this paper, we present the results of applying different, recently developed methods to improve deep neural network training and operating. We decided to focus on the most popular CNN structures, namely on VGG based neural networks: VGG16, VGG11 and proposed by us VGG8. The tests were conducted on a real and very important problem of skin cancer detection. A publicly available dataset of skin lesions was used as a benchmark. We analyzed the influence of applying: dropout, batch normalization, model ensembling, and transfer learning. Moreover, the influence of the type of activation function was checked. In order to increase the objectivity of the results, each of the tested models was trained 6 times and their results were averaged. In addition, in order to mitigate the impact of the selection of learning, test and validation sets, k-fold validation was applied.

A detailed comparison of catalytic properties of two different ruthenium-based catalysts in the reaction of homogeneous hydrogenation of acetophenone was performed. Additionally, methods of synthesis of both catalysts were tested and optimized in order to achieve the best possible quality and purity of the final catalysts.

NMR analysis was used to analyze and identify the composition of ruthenium compounds and gas chromatography was used to analyze the conversion rate of hydrogenation reactions.

It was determined that RuCl2(PPh3)3 obtained with a modified method described by Shaw’s group (Shawet al., 2007) had the best catalytic properties in the reaction performed under conditions described in Liang Wang’s publication (Wang et al., 2014). It was also determined that for concentration ratio of substrate to RuCl2(PPh3)3 amounting to 250:1 the conversion rate was much higher than that of the reaction performed with a double dose of the catalyst. Results of experiments also show that samples of the post-reaction solution should be analyzed right after the reaction, because even if they are stored in low temperature the amount of product can change up to 3–5% compared to the base sample and this change is not predictable.

These findings have significant implications for further research of the reaction of homogeneous transfer hydrogenation of ketones. With the right catalysts and methods of their synthesis other parameters of this reaction can be optimized. The most important one is a change of solvent from isopropyl alcohol to a less toxic substance like water. This may increase the value of the reaction in green chemistry and chemical industry.

Malignant melanomas are the most deadly type of skin cancer, yet detected early have high chances of successful treatment. In the last twenty years, the interest in automatic recognition and classification of melanoma dynamically increased, partly because of appearing public datasets with dermatoscopic images of skin lesions. Automated computer-aided skin cancer detection in dermatoscopic images is a very challenging task due to uneven sizes of datasets, huge intra-class variation with small interclass variation, and the existence of many artifacts in the images. One of the most recognized methods of melanoma diagnosis is the ABCD method. In the paper, we propose an extended version of this method and an intelligent decision support system based on neural networks that uses its results in the form of hand-crafted features. Automatic determination of the skin features with the ABCD method is difficult due to the large diversity of images of various quality, the existence of hair, different markers and other obstacles. Therefore, it was necessary to apply advanced methods of pre-processing the images. The proposed system is an ensemble of ten neural networks working in parallel, and one network using their results to generate a final decision. This system structure enables to increase the efficiency of its operation by several percentage points compared with a single neural network. The proposed system is trained on over 5000 and tested afterwards on 200 skin moles. The presented system can be used as a decision support system for primary care physicians, as a system capable of self-examination of the skin with a dermatoscope and also as an important tool to improve biopsy decision making.

In this paper the authors propose a decision support system for automatic blood smear analysis based on microscopic images. The images are pre-processed in order to remove irrelevant elements and to enhance the most important ones – the healthy blood cells (erythrocytes) and the pathologic ones (echinocytes). The separated blood cells are analysed in terms of their most important features by the eigenfaces method. The features are the basis for designing the neural network classifier, learned to distinguish between erythrocytes and echinocytes. As the result, the proposed system is able to analyse the smear blood images in a fully automatic way and to deliver information on the number and statistics of the red blood cells, both healthy and pathologic. The system was examined in two case studies, involving the canine and human blood, and then consulted with the experienced medicine specialists. The accuracy of classification of red blood cells into erythrocytes and echinocytes reaches 96%.

The paper presents verification of a peak detection method cooperating with infrared radiation detector module applications. The work has been divided into parts including SPICE simulations and presentation of results obtained with the constructed prototype. The design of the peak detector dedicated to applications with very short pulses requires a different approach than that for standard solutions. It is mainly caused due to the ratio of pulse width and time period. In the described application this ratio is less than 10%. The paper shows testing of an analogue circuit which is capable to be inserted in these applications.

Due to the unrecognized effect of tigecycline (TIG) on CD4+ and CD8+ T cells, the present study has been undertaken in order to determine whether the drug can affect these cells in respect of their counts, and the production of IFN-γ, IL-17 (pro-inflammatory and immune-protective cytokines), IL-4 (anti-inflammatory and immune-protective cytokine), IL-10 and TGF-β (anti-inflammatory and immune-suppressive cytokines). Murine lymphocytes were treated with TIG for 48 and 96 h at concentrations reflecting its plasma levels obtained in vivo at therapeutic doses, and at 10-fold lower concentrations. It was found that TIG neither affected substantially the percentage and absolute counts of entire CD4+ and CD8+ T cell populations nor influenced the Foxp3+CD25+CD4+ regulatory/suppressive T cell subset. Furthermore, the percentages of IL-4-, IL-10-, IL-17- and TGF-β-producing CD4+ T cells were not altered following the exposure to TIG. Similarly, TIG did not influence IFN-γ production by CD8+ T cells. Thus, with respect to the parameters evaluated, TIG does not seem to exert immune-suppressive and anti-inflammatory effects.

The article presents state of work in technology of free-space optical communications (Free Space Optics − FSO). Both commercially available optical data links and their further development are described. The main elements and operation limiting factors of FSO systems have been identified. Additionally, analyses of FSO/RF hybrid systems application are included. The main aspects of LasBITer project related to such hybrid technology for security and defence applications are presented.

Detection of explosives vapors is an extremely difficult task. The sensitivity of currently constructed detectors is often insufficient. The paper presents a description of an explosive vapors concentrator that improves the detection limit of some explosives detectors. These detectors have been developed at the Institute of Optoelectronics. The concentrator is especially dedicated to operate with nitrogen oxide detectors. Preliminary measurements show that using the concentrator, the recorded amount of nitrogen dioxide released from a 0.5 ng sample of TNT increases by a factor of approx. 20. In the concentrator an induction heater is applied. Thanks to this and because of the miniaturization of the container with an adsorbing material (approx. 1 cm3), an extremely high rate of temperature growth is achieved (up to 500 °C within approx. 25 s). The concentration process is controlled by a microcontroller. Compact construction and battery power supply provide a possibility of using the concentrator as a portable device.

The aim of this study was to determine the influence of feed on the pharmacokinetics of flumequine (FLU) administered to broiler chickens as follows: directly into the crop (10 mg/kg of BW) of fasted (group I/control) and non-fasted chickens (group II), or administered continu- ously with drinking water (1 g/L for 72 h) and with unlimited access to feed (group III). Plasma concentration of FLU was determined by high-performance liquid chromatography with fluo- rescence detection. In group II, a significant decrease in the maximum concentration (Cmax = 2.13±0.7 μg/mL) and the area under the concentration curve from zero to infinity (AUC0→∞ = 7.47±2.41 μg·h/mL) was noted as compared to the control group (Cmax = 4.11±1.68 μg/mL and AUC0→∞ = 18.17±6.85 μg·h/mL, respectively). In group III, the decrease in AUC was signifi- cant only in the first 3 hours (AUC0→3 = 5.02±1.34 μg·h/mL) as compared to the control group (AUC0→3 = 7.79±3.29 μg·h/mL). The results indicate that feed reduced the bioavailability of FLU from the gastrointestinal tract by at least 50% after the administration of a single oral dose. However, continuous administration of FLU with drinking water could compensate for the feed-induced decrease in absorption after single oral dose.