This paper presents investigations on the removal of cyclohexane and ethanol from air in polyurethane- -packed biotrickling filters, inoculated with Candida albicans and Candida subhashii fungal species. Results on process performance together with flow cytometry analyses of the biofilm formed over packing elements are presented and discussed. The results indicate that the presence of ethanol enhances the removal efficiency of cyclohexane from air. This synergistic effect may be attributed to both co-metabolism of cyclohexane with ethanol as well as increased sorption efficiency of cyclohexane to mineral salt medium in the presence of ethanol. Maximum elimination capacities of 89 g m-3 h-1 and 36.7 g m-3 h-1 were noted for cyclohexane and ethanol, respectively, when a mixture of these compounds was treated in a biofilter inoculated with C. subhashii. Results of flow cytometry analyses after 100 days of biofiltration revealed that about 91% and 88% of cells in biofilm remained actively dividing, respectively for C. albicans and C. subhashii species, indicating their good condition and ability to utilize cyclohexane and ethanol as a carbon source.

The paper presents the fusion approach of different feature selection methods in pattern recognition problems. The following methods are examined: nearest component analysis, Fisher discriminant criterion, refiefF method, stepwise fit, Kolmogorov-Smirnov criteria, T2-test, Kruskall-Wallis test, feature correlation with class, and SVM recursive feature elimination. The sensitivity to the noisy data as well as the repeatability of the most important features are studied. Based on this study, the best selection methods are chosen and applied in the process of selection of the most important genes and gene sequences in a dataset of gene expression microarray in prostate and ovarian cancers. The results of their fusion are presented and discussed. The small selected set of such genes can be treated as biomarkers of cancer.

For brain tumour treatment plans, the diagnoses and predictions made by medical doctors and radiologists are dependent on medical imaging. Obtaining clinically meaningful information from various imaging modalities such as computerized tomography (CT), positron emission tomography (PET) and magnetic resonance (MR) scans are the core methods in software and advanced screening utilized by radiologists. In this paper, a universal and complex framework for two parts of the dose control process – tumours detection and tumours area segmentation from medical images is introduced. The framework formed the implementation of methods to detect glioma tumour from CT and PET scans. Two deep learning pre-trained models: VGG19 and VGG19-BN were investigated and utilized to fuse CT and PET examinations results. Mask R-CNN (region-based convolutional neural network) was used for tumour detection – output of the model is bounding box coordinates for each object in the image – tumour. U-Net was used to perform semantic segmentation – segment malignant cells and tumour area. Transfer learning technique was used to increase the accuracy of models while having a limited collection of the dataset. Data augmentation methods were applied to generate and increase the number of training samples. The implemented framework can be utilized for other use-cases that combine object detection and area segmentation from grayscale and RGB images, especially to shape computer-aided diagnosis (CADx) and computer-aided detection (CADe) systems in the healthcare industry to facilitate and assist doctors and medical care providers.

Recently, the analysis of medical imaging is gaining substantial research interest, due to advancements in the computer vision field. Automation of medical image analysis can significantly improve the diagnosis process and lead to better prioritization of patients waiting for medical consultation. This research is dedicated to building a multi-feature ensemble model which associates two independent methods of image description: textural features and deep learning. Different algorithms of classification were applied to single-phase computed tomography images containing 8 subtypes of renal neoplastic lesions. The final ensemble includes a textural description combined with a support vector machine and various configurations of Convolutional Neural Networks. Results of experimental tests have proved that such a model can achieve 93.6% of weighted F1-score (tested in 10-fold cross validation mode). Improvement of performance of the best individual predictor totalled 3.5 percentage points.

The article presents research on the use of Monte-Carlo Tree Search (MCTS) methods to create an artificial player for the popular card game “The Lord of the Rings”. The game is characterized by complicated rules, multi-stage round construction, and a high level of randomness. The described study found that the best probability of a win is received for a strategy combining expert knowledge-based agents with MCTS agents at different decision stages. It is also beneficial to replace random playouts with playouts using expert knowledge. The results of the final experiments indicate that the relative effectiveness of the developed solution grows as the difficulty of the game increases.

The paper is focused on automatic segmentation task of bone structures out of CT data series of pelvic region. The authors trained and compared four different models of deep neural networks (FCN, PSPNet, U-net and Segnet) to perform the segmentation task of three following classes: background, patient outline and bones. The mean and class-wise Intersection over Union (IoU), Dice coefficient and pixel accuracy measures were evaluated for each network outcome. In the initial phase all of the networks were trained for 10 epochs. The most exact segmentation results were obtained with the use of U-net model, with mean IoU value equal to 93.2%. The results where further outperformed with the U-net model modification with ResNet50 model used as the encoder, trained by 30 epochs, which obtained following result: mIoU measure – 96.92%, “bone” class IoU – 92.87%, mDice coefficient – 98.41%, mDice coefficient for “bone” – 96.31%, mAccuracy – 99.85% and Accuracy for “bone” class – 99.92%.

In this article, an analytical-numerical approach to calculating a stationary thermal field in the elliptical region is presented. The eigenfunctions of the Laplace operator were determined analytically, whereas the coefficients of the eigenfunctions were obtained numerically. The cooling was modeled with 3rd kind (Hankel’s) boundary condition, where the total heat transfer coefficient was the sum of the convective and radiative components. The method was used to analyze the thermal field in an elliptical conductor and a dielectrically heated elliptical column. The basic parameters of these systems, i.e. their steady-state current rating and the maximum charge temperature, were determined. The results were verified using the finite element method and have been presented graphically.

The paper features a grid-tied converter with a repetitive current controller. Our goal here is to demonstrate the complete design workflow for a repetitive controller, including phase lead, filtering and conditional learning. All key parameters, i.e., controller gain, filter and fractional phase lead, are designed in a single optimization procedure, which is a novel approach. The description of the design and optimization process, as well as experimental verification of the entire control system, are the most important contributions of the paper. Additionally, one more novelty in the context of power converters is verified in the physical system – a conditional learning algorithm to improve transient states to abrupt reference and disturbance changes. The resulting control system is tested experimentally in a 10 kW converter.