Hybrid Block-Based Lightweight Machine Learning-Based Predictive Models for Quality Preserving in the Internet of Things- (IoT-) Based Medical Images with Diagnostic Applications.

In the contemporary era of unprecedented innovations such as the Internet of Things (IoT), modern applications cannot be imagined without the presence of a wireless sensor network (WSN). Nodes in WSN use neighbor discovery (ND) protocols to have necessary communication among the nodes. The neighbor discovery process is crucial as it is to be done with energy efficiency and minimize discovery latency and maximum percentage of neighbors discovered. The current ND approaches that are indirect in nature are categorized into methods of removal of active slots from wake-up schedules and intelligent addition of new slots. This work develops a lightweight intrusion detection system (IDS) based on two machine learning approaches, namely, feature selection and feature classification, in order to improve the security of the Internet of Things (IoT) while transferring medical data through a cloud platform. In order to take advantage of the comparatively cheap processing cost of the filter-based technique, the feature selection was carried out. The two methods are found to have certain drawbacks. The first category disturbs the original integrity of wake-up schedules leading to reduced chances of discovering new nodes in WSN as neighbors. When the second category is followed, it may have inefficient slots in the wake-up schedules leading to performance degradation. Therefore, the motivation behind the work in this paper is that by combining the two categories, it is possible to reap the benefits of both and get rid of the limitations of both. Making a hybrid is achieved by introducing virtual nodes that help maximize performance by ensuring the original integrity of wake-up schedules and adding efficient active slots. Thus, a Hybrid Approach to Neighbor Discovery (HAND) protocol is realized in WSN. The simulation study revealed that HAND outperforms the existing indirect ND models.

Detection of Breast Cancer Using Histopathological Image Classification Dataset with Deep Learning Techniques.

Cancer is one of the top causes of mortality, and it arises when cells in the body grow abnormally, like in the case of breast cancer. For people all around the world, it has now become a huge issue and a threat to their safety and wellbeing. Breast cancer is one of the major causes of death among females all over the globe, and it is particularly prevalent in the United States. It is possible to diagnose breast cancer using a variety of imaging modalities including mammography, computerized tomography (CT), magnetic resonance imaging (MRI), ultrasound, and biopsies, among others. To analyze the picture, a histopathology study (biopsy) is often performed, which assists in the diagnosis of breast cancer. The goal of this study is to develop improved strategies for various CAD phases that will play a critical role in minimizing the variability gap between and among observers. It created an automatic segmentation approach that is then followed by self-driven post-processing activities to successfully identify the Fourier Transform based Segmentation in the CAD system to improve its performance. When compared to existing techniques, the proposed segmentation technique has several advantages: spatial information is incorporated, there is no need to set any initial parameters beforehand, it is independent of magnification, it automatically determines the inputs for morphological operations to enhance segmented images so that pathologists can analyze the image with greater clarity, and it is fast. Extensive tests were conducted to determine the most effective feature extraction techniques and to investigate how textural, morphological, and graph characteristics impact the accuracy of categorization classification. In addition, a classification strategy for breast cancer detection has been developed that is based on weighted feature selection and uses an upgraded version of the Genetic Algorithm in conjunction with a Convolutional Neural Network Classifier. The practical application of the suggested improved segmentation and classification algorithms for the CAD framework may reduce the number of incorrect diagnoses and increase the accuracy of classification. So, it may serve as a second opinion tool for pathologists and aid in the early detection of diseases.