University of Central Asia

Plant diseases represent a persistent challenge to global food security, with annual crop yield losses reaching as high as 40%. Timely and reliable identification of these diseases is therefore essential for reducing economic losses and ensuring sustainable farming practices. Although deep learning has achieved remarkable success in image-based disease classification, much of the existing research emphasizes complex convolutional neural networks that are unsuitable for resource-constrained environments. In this work, we present a systematic evaluation of two widely adopted transfer learning models ResNet50 and MobileNetV2 applied to the New Plant Disease dataset (~87,000 images spanning 38 categories). Our experiments examine the role of augmentation, learning rates, batch sizes, and fine-tuning strategies in model performance. Results demonstrate that MobileNetV2 not only achieved superior accuracy (98.33%) with strong precision, recall, and AUC values, but also required seven times fewer parameters and significantly reduced training time compared with ResNet50. Error analysis further revealed MobileNetV2’s ability to differentiate between diseases with overlapping symptoms. Importantly, its lightweight architecture supports real-time implementation on mobile devices, drones, and IoT systems, offering clear advantages for field deployment. Unlike prior studies that emphasize raw accuracy, this research highlights efficiency, robustness, and deployability. Overall, our findings establish MobileNetV2 as a practical and scalable solution for next-generation precision agriculture and low-cost disease monitoring systems.

Uremia has been associated with slowed intestinal motility, resulting in loss of peristalsis and the development of paralytic ileus. We present a case of uremia-induced paralytic ileus in a 70-year-old woman. The patient was admitted with a 7-day history of abdominal pain, vomiting, and constipation. Initial evaluation revealed a soft non-tender abdomen with decreased bowel sounds. Urgent ultrasonography showed dilated gut loops and sluggish gut motility, while atrophic kidneys with hydronephrosis were also appreciated. Routine biochemistry labs revealed elevated creatinine and potassium levels, and an erect abdominal X-ray showed mild to moderate air-fluid levels. Notably, common risk factors for paralytic ileus, such as recent surgery or gastrointestinal procedures, immobility, hypokalemia, hypothyroidism, hypercalcaemia, hypocalcemia, hypomagnesemia, and diabetes mellitus, were ruled out. The patient also had no history of anticholinergic or opioid use. Given the absence of other causes, uremia was established as the underlying cause of paralytic ileus. After initial treatment for ileus, the patient underwent hemodialysis, which led to significant improvement in symptoms. This case highlights the importance of recognizing uremia as a potential cause of unexplained ileus, contributing to the limited literature on uremia-associated ileus.

In this study, a novel GO/Fe@TANG composite catalyst was designed by functionalizing graphene oxide (GO) with iron (Fe) and incorporating a hydrogen (H 2 ) trapping TANG COF.

The importance of integration and retrieval of data is growing more and more in the context of post-COVID-19 analysis because of the increased generation of data in research and resources for studying COVID-19. In this context, the analysis of contagion groups that share some specific feature (for example, people that work together, people with the same clinical manifestation, etc.) may offer remarkable insights about COVID-19. For instance, it can be helpful to recognize the behavior of this disease in accordance with the unique characteristics of a group of people. In this regard, ontologies are a widely accepted alternative for representing and analyzing knowledge. Given their benefits, this paper introduces an ontology-based method to formally describe and analyze information about specific groups of people with COVID-19. Since the ontology was specified in OWL, a formal language based on description logics, it enables the consistency of the represented information to be verified and use automatic reasoning to deduce new knowledge. This approach allows modeling a wide variety of characteristics (symptoms, comorbidities, treatments, etc.) of the individuals and consequently use it to infer the general characteristics of the groups that they belong to. Hence, a reasoner can be applied to perform advanced analysis either to identify patterns in these groups or to find similarity with other groups. To demonstrate the applicability of this method, a case study is described. In addition, the ontology was used to represent and analyze the information of a sample of patients extracted from a public dataset. The results demonstrate the capability of the ontology to represent and analyze information about specific groups of people with COVID-19.

Mucormycosis is a rare opportunistic fungal infection caused by mucormycetes, primarily affecting immunocompromised individuals such as diabetic patients. Cranial nerve involvement, including facial and abducens nerve palsies, is an uncommon but serious complication. We present the case of a 47-year-old female with uncontrolled diabetes and iatrogenic Cushing's syndrome who developed mucormycosis complicated by right-sided facial weakness (upper motor neuron type) and abducens nerve palsy (lower motor neuron type). She presented with a 1-month history of right-sided facial swelling, headache, and earache, and a 6-day history of sudden-onset facial weakness. Examination revealed preserved eye closure and forehead wrinkling, with a deficit in right eye abduction. Black eschar was noted in the intranasal and hard palate areas. Imaging (computed tomography paranasal sinuses and magnetic resonance imaging) and histopathological confirmation established the diagnosis. Treatment included amphotericin B and 2 surgical debridements following otorhinolaryngology consultation. Patient's facial swelling improved during hospital stay, and abduction deficit resolved on 1-month follow-up. This case highlights the critical need for early recognition and multidisciplinary management of mucormycosis, particularly in diabetic or immunocompromised patients presenting with cranial nerve deficits. Prompt diagnosis and treatment are vital to reduce morbidity and improve outcomes.

In cellular biology and regenerative medicine, the application of deep learning becomes a groundbreaking technology, meanwhile using CNNs for biological information. Among the sample types, eosinophils account for 3,117 samples while platelets total 2,348 samples in the current dataset. The deep learning model gets a training accuracy of 98.5 per cent, a test accuracy is 97.5 per cent, and the test loss is equal to 0.7024. Stem cell differentiation, and personalized regenerative medicine, should benefit from developments in multi-omics data analysis. Thus, some concerns enhance deep learning, yet offer promise for innovations in precision medicine and biological study.