Extended Spectrum beta-Lactamase Bacteria and Multidrug Resistance in Jordan are Predicted Using a New Machine-Learning system.

Background: The incidence of microorganisms with extended-spectrum beta-lactamase (ESBL) is on the rise, posing a significant public health concern. The current application of machine learning (ML) focuses on predicting bacterial resistance to optimize antibiotic therapy. This study employs ML to forecast the occurrence of bacteria that generate ESBL and demonstrate resistance to multiple antibiotics (MDR). Methods: Six popular ML algorithms were initially trained on antibiotic resistance test patient reports (n = 489) collected from Al-Hussein/Salt Hospital in Jordan. Trained outcome models predict ESBL and multidrug resistance profiles based on microbiological and patients' clinical data. The results were utilized to select the optimal ML method to predict ESBL's most associated features. Results: Escherichia coli (E. coli, 82%) was the most commonly identified microbe generating ESBL, displaying multidrug resistance. Urinary tract infections (UTIs) constituted the most frequently observed clinical diagnosis (68.7%). Classification and Regression Trees (CART) and Random Forest (RF) classifiers emerged as the most effective algorithms. The relevant features associated with the emergence of ESBL include age and different classes of antibiotics, including cefuroxime, ceftazidime, cefepime, trimethoprim/ sulfamethoxazole, ciprofloxacin, and gentamicin. Fosfomycin nitrofurantoin, piperacillin/tazobactam, along with amikacin, meropenem, and imipenem, had a pronounced inverse relationship with the ESBL class. Conclusion: CART and RF-based ML algorithms can be employed to predict the most important features of ESBL. The significance of monitoring trends in ESBL infections is emphasized to facilitate the administration of appropriate antibiotic therapy.

Multiple Mobile Sinks for Quality of Service Improvement in Large-Scale Wireless Sensor Networks.

The involvement of wireless sensor networks in large-scale real-time applications is exponentially growing. These applications can range from hazardous area supervision to military applications. In such critical contexts, the simultaneous improvement of the quality of service and the network lifetime represents a big challenge. To meet these requirements, using multiple mobile sinks can be a key solution to accommodate the variations that may affect the network. Recent studies were based on predefined mobility models for sinks and relied on multi-hop routing techniques. Besides, most of these studies focused only on improving energy consumption without considering QoS metrics. In this paper, multiple mobile sinks with random mobile models are used to establish a tradeoff between power consumption and the quality of service. The simulation results show that using hierarchical data routing with random mobile sinks represents an efficient method to balance the distribution of the energy levels of nodes and to reduce the overall power consumption. Moreover, it is proven that the proposed routing methods allow for minimizing the latency of the transmitted data, increasing the reliability, and improving the throughput of the received data compared to recent works, which are based on predefined trajectories of mobile sinks and multi-hop architectures.

COVID-19 Diagnosis and Classification Using Radiological Imaging and Deep Learning Techniques: A Comparative Study.

In December 2019, the novel coronavirus disease 2019 (COVID-19) appeared. Being highly contagious and with no effective treatment available, the only solution was to detect and isolate infected patients to further break the chain of infection. The shortage of test kits and other drawbacks of lab tests motivated researchers to build an automated diagnosis system using chest X-rays and CT scanning. The reviewed works in this study use AI coupled with the radiological image processing of raw chest X-rays and CT images to train various CNN models. They use transfer learning and numerous types of binary and multi-class classifications. The models are trained and validated on several datasets, the attributes of which are also discussed. The obtained results of various algorithms are later compared using performance metrics such as accuracy, F1 score, and AUC. Major challenges faced in this research domain are the limited availability of COVID image data and the high accuracy of the prediction of the severity of patients using deep learning compared to well-known methods of COVID-19 detection such as PCR tests. These automated detection systems using CXR technology are reliable enough to help radiologists in the initial screening and in the immediate diagnosis of infected individuals. They are preferred because of their low cost, availability, and fast results.