Dengue endemicity has become regular in recent times across the world. The numbers of cases and deaths have been alarmingly increasing over the years. In addition to this, there are no direct medications or vaccines to treat this viral infection. Thus, monitoring and controlling the carriers of this virus which are the Aedes mosquitoes become specially demanding to combat the endemicity, as killing all the mosquitoes regardless of their species would destroy ecosystems. The current approach requires collecting a larva sample from the hatching sites and, then, an expert entomologist manually examining it using a microscope in the laboratory to identify the Aedes vector. This is time-consuming, labor-intensive, subjective, and impractical. Several automated Aedes larvae detection systems have been proposed previously, but failed to achieve sufficient accuracy and reliability. We propose an automated system utilizing ensemble learning, which detects Aedes larvae effectively from a low-magnification image with an accuracy of over 99%. The proposed system outperformed all the previous methods with respect to accuracy. The practical usability of the system is also demonstrated.

Among the prevalent cyberattacks on Android devices, a ransomware attack is the most common and damaging. Although there are many solutions for detecting Android ransomware attacks, existing solutions have limited detection accuracy and high computational complexity. This paper proposes a new Android ransomware detection method based on traffic analysis to address the limitations. We exploit particle swarm optimization (PSO) to select traffic characteristics. Then, based on the selected traffic features, we classify the data traffic using decision tree and random forest classifiers. We examine ransomware cyberattacks in two distinct circumstances. In the first case, we find ransomware traffic; in the second, we locate a specific form of malware traffic among benign traffic. The proposed PSO-assisted feature selection enables the classifier to improve the detection accuracy significantly. The random forest is found to achieve the highest performance in detecting ransomware, whereas the decision tree is the best for detecting the types of ransomware. The accuracy improvements are 2.26% and 3.7% in the first and second scenarios, respectively. The proposed method removes 56.01% to 91.95% of the features. The proposed method convergences quickly, as the optimization reaches an optimum value of about ten iterations.

Traditionally, pathological analysis and diagnosis are performed by manually eyeballing glass-slide specimens under a microscope by an expert. The whole slide image (WSI) is the digital specimen produced from the glass slide. WSI enabled specimens to be observed on a computer screen and led to computational pathology where computer vision and artificial intelligence are utilized for automated analysis and diagnosis. With the current computational advancement, the entire WSI can be analyzed autonomously without human supervision. However, the analysis could fail or lead to wrong diagnosis if the WSI is affected by tissue artifacts such as tissue fold or air bubbles depending on the severity. Existing artifact detection methods rely on experts for severity assessment to eliminate artifact-affected regions from the analysis. This process is time-consuming, exhausting and undermines the goal of automated analysis or removal of artifacts without evaluating their severity, which could result in the loss of diagnostically important data. Therefore, it is necessary to detect artifacts and then assess their severity automatically. In this paper, we propose a system that incorporates severity evaluation with artifact detection utilizing convolutional neural networks (CNN). The proposed system uses DoubleUNet to segment artifacts and an ensemble network of six fine-tuned CNN models to determine severity. This method outperformed current state-of-the-art in accuracy by 9% for artifact segmentation and achieved a strong correlation of 97% with the pathologist’s evaluation for severity assessment. The robustness of the system was demonstrated using our proposed heterogeneous dataset and practical usability was ensured by integrating it with an automated analysis system.

Flying base stations (FlyBSs) enable ubiquitous communications in the next generation mobile networks with a flexible topology. However, a deployment of the FlyBSs intensifies interference, which can result in a degradation in the throughput of cell-edge users. In this paper, we introduce a flexible soft frequency reuse (F-SFR) that enables a self-organization of a common SFR in the networks with an unpredictable and dynamic topology with the FlyBSs. We propose a graph theory-based algorithm for an allocation of resource plans, which is understood as a bandwidth allocation and a transmission power setting in the context of SFR. Furthermore, we introduce a low-complexity implementation of the proposed resource allocation using deep neural network (DNN) to significantly reduce the computation complexity. We show that the proposed F-SFR increases the throughput of cell-edge users by 16% to 26% and, at the same time, improves the satisfaction of the cell-edge users by up to 25% compared to the state-of-the-art solutions. We also demonstrate that the proposed scheme ensures a higher fairness in the throughput among the users with respect to the state-of-the-art solutions. The implementation via DNN also outperforms all state-of-the-art solutions despite its very low complexity.

In this letter, we propose a low-complexity data-null subcarrier switching-based peak-to-average power ratio (PAPR) reduction scheme for the orthogonal frequency division multiplexing systems, which provides improved bit error rate. We perform the switching between the data and null subcarriers in such a way that distance between any two switched null subcarriers remains constant. This method can achieve up to 6 dB signal-to-noise ratio gain and shed the system complexity by more than 98% of the conventional scheme, with a slight compromise of the PAPR reduction capability.

Human epidermal growth factor receptor 2 (HER2) quantification is performed routinely for all breast cancer patients to determine their suitability for HER2-targeted therapy. Fluorescence in situ hybridization (FISH) and chromogenic in situ hybridization (CISH) are the US Food and Drug Administration (FDA) approved tests for HER2 quantification in which at least 20 cancer-affected singular nuclei are quantified for HER2 grading. CISH is more advantageous than FISH for cost, time and practical usability. In clinical practice, nuclei suitable for HER2 quantification are selected manually by pathologists which is time-consuming and laborious. Previously, a method was proposed for automatic HER2 quantification using a support vector machine (SVM) to detect suitable singular nuclei from CISH slides. However, the SVM-based method occasionally failed to detect singular nuclei resulting in inaccurate results. Therefore, it is necessary to develop a robust nuclei detection method for reliable automatic HER2 quantification. In this paper, we propose a robust U-net-based singular nuclei detection method with complementary color correction and deconvolution adapted for accurate HER2 grading using CISH whole slide images (WSIs). The efficacy of the proposed method was demonstrated for automatic HER2 quantification during a comparison with the SVM-based approach.