The prevalence and impact of university affiliation discrepancies between four bibliographic databases – Scopus, Web of Science, Dimensions, and Microsoft Academic

Research managers benchmarking universities against international peers face the problem of affiliation disambiguation. Different databases have taken separate approaches to this problem and discrepancies exist between them. Bibliometric data sources typically conduct a disambiguation process that unifies variant institutional names and those of its sub-units so that researchers can then search all records from that institution using a single unified name. This study examined affiliation discrepancies between Scopus, Web of Science, Dimensions, and Microsoft Academic for 18 Arab universities over a five-year period. We confirmed that digital object identifiers (DOIs) are suitable for extracting comparable scholarly material across databases and quantified the affiliation discrepancies between them. A substantial share of records assigned to the selected universities in any one database were not assigned to the same university in another. The share of discrepancy was higher in the larger databases, Dimensions and Microsoft Academic. The smaller, more selective databases, Scopus and especially Web of Science tended to agree to a greater degree with affiliations in the other databases. Manual examination of affiliation discrepancies showed they were caused by a mixture of missing affiliations, unification differences, and assignation of records to the wrong institution. Peer Review https://publons.com/publon/10.1162/qss_a_00175

Plant disease prediction using convolutional neural network

Every year India losses the significant amount of annual crop yield due to unidentified plant diseases. The traditional method of disease detection is manual examination by either farmers or experts, which may be time-consuming and inaccurate. It is proving infeasible for many small and medium-sized farms around the world. To mitigate this issue, computer aided disease recognition model is proposed. It uses leaf image classification with the help of deep convolutional networks. In this paper, VGG16 and Resnet34 CNN was proposed to detect the plant disease. It has three processing steps namely feature extraction, downsizing image and classification. In CNN, the convolutional layer extracts the feature from plant image. The pooling layer downsizing the image. The disease classification was done in dense layer. The proposed model can recognize 38 differing types of plant diseases out of 14 different plants with the power to differentiate plant leaves from their surroundings. The performance of VGG16 and Resnet34 was compared. The accuracy, sensitivity and specificity was taken as performance Metrix. It helps to give personalized recommendations to the farmers based on soil features, temperature and humidity

Income level and antibiotic misuse: a systematic review and dose–response meta-analysis

Objectives To quantify the association between income and antibiotic misuse including unprescribed use, storage of antibiotics and non-adherence. Methods We identified pertinent studies through database search, and manual examination of reference lists of selected articles and review reports. We performed a dose–response meta-analysis of income, both continuous and categorical, in relation to antibiotic misuse. Summary odds ratios (ORs) and their 95% confidence intervals (CIs) were estimated under a random-effects random effects model. Results Fifty-seven studies from 22 countries of different economic class were included. Overall, the data are in agreement with a flat linear association between income standardized to socio-economic indicators and antibiotic misuse (OR per 1 unit increment = 1.00, p-value = 0.954, p-value non-linearity = 0.429). Data were compatible with no association between medium and high income with general antibiotic misuse (OR 1.04; 95% CI 0.89, 1.20 and OR 1.03; 95% CI 0.82, 1.29). Medium income was associated with 19% higher odds of antibiotic storage (OR 1.19; 95% CI 1.07, 1.32) and 18% higher odds of any aspect of antibiotic misuse in African studies (OR 1.18; 95% CI 1.00, 1.39). High income was associated with 51% lower odds of non-adherence to antibiotic treatment (OR 0.49; 95% CI 0.34, 0.60). High income was also associated with 11% higher odds of any antibiotic misuse in upper-middle wealth countries (OR 1.11; 95% CI 1.00, 1.22). Conclusions The association between income and antibiotic misuse varies by type of misuse and country wellness. Understanding the socioeconomic properties of antibiotic misuse should prove useful in developing related intervention programs and health policies.

Pancreatic cancer grading in pathological images using deep learning convolutional neural networks

Background: Pancreatic cancer is one of the deadliest forms of cancer. The cancer grades define how aggressively the cancer will spread and give indication for doctors to make proper prognosis and treatment. The current method of pancreatic cancer grading, by means of manual examination of the cancerous tissue following a biopsy, is time consuming and often results in misdiagnosis and thus incorrect treatment. This paper presents an automated grading system for pancreatic cancer from pathology images developed by comparing deep learning models on two different pathological stains. Methods: A transfer-learning technique was adopted by testing the method on 14 different ImageNet pre-trained models. The models were fine-tuned to be trained with our dataset. Results: From the experiment, DenseNet models appeared to be the best at classifying the validation set with up to 95.61% accuracy in grading pancreatic cancer despite the small sample set. Conclusions: To the best of our knowledge, this is the first work in grading pancreatic cancer based on pathology images. Previous works have either focused only on detection (benign or malignant), or on radiology images (computerized tomography [CT], magnetic resonance imaging [MRI] etc.). The proposed system can be very useful to pathologists in facilitating an automated or semi-automated cancer grading system, which can address the problems found in manual grading.

Clinical, endoscopic and radiological assessment of patients with obstructed defecation: findings and management at a District General Hospital

Background: The aim of the study was to determine the diagnostic value of clinical, endoscopic and proctographic assessment as well as clinical outcomes in patients with obstructed defaecation (OD). The study also examined correlation between clinical/endoscopic findings and proctogram in the diagnosis of rectocele and intra-rectal intussusception (IRI).Methods: Patients presenting with symptoms of OD between January-December 2018 were assessed with manual examination, endoscopy and defecation proctogram. Patients were followed for 2-3 years for clinical outcomes.Results: There were 65 female (97.01%) and 2 male patients (2.98%), with an average age of 57.77 (34-88) years. Main indications were OD, altered bowels, faecal urgency and rectal bleeding. A total of 67 X-ray defecating proctograms and 77 endoscopies were performed. Main findings on clinico-endoscopic examination were IRI (44), rectocele (36) and haemorrhoids (21). Main findings on proctogram were rectocele (59), IRI (56) and enterocele (13). Endoscopic assessment showed sensitivity: 55.93%, specificity: 62.50% and accuracy: 56.72% in diagnosing rectocele when compared with the diagnostic confirmation on proctogram. Combining manual assessment with endoscopic findings improved sensitivity (76.27%) and accuracy (68.66%). Similar improvement was also noted in the sensitivity (61.40 to 66.67%), specificity (47 to 58%), and accuracy (53.73 to 58.21%) in diagnosing IRI when compared with the diagnostic confirmation on proctogram. Majority of the patients improved with conservative measures; however, surgical intervention was required in 13 patients.Conclusions: Although manual examination enhances endoscopic assessment in diagnosing rectocele and IRI, proctogram is still required for objective assessment. Management of OD remains mainly conservative, with surgical intervention required in some patients.

Asbestos bodies count and morphometry in bulk lung tissue samples by non-invasive X-ray micro-tomography

The number of the Asbestos Bodies (AB), i.e. asbestos that developed an iron-protein coating during its permanence in biological tissues, is one of the most accessible markers of asbestos exposure in individuals. The approaches developed to perform AB count in biological tissues are based on the manual examination of tissue digests or histological sections by means of light or electron microscopies. Although these approaches are well established and relatively accessible, manual examination is time-consuming and can be reader-dependent. Besides, approximations are applied because of the limitations of 2D readings and to speed up manual counts. In addition, sample preparation using tissue digests require an amount of tissue that can only be obtained by invasive surgery or post-mortem sampling. In this paper, we propose a new approach to AB counting based on non-destructive 3D imaging, which has the potential to overcome most of the limitations of conventional approaches. This method allows automating the AB count and determining their morphometry distribution in bulk tissue samples (ideally non-invasive needle biopsies), with minimal sample preparation and avoiding approximations. Although the results are promising, additional testing on a larger number of AB-containing biological samples would be required to fully validate the method.

Intelligent Fluorescence Image Analysis of Giant Unilamellar Vesicles and Protein Liquid Phase Droplets by Whole Z-stack Analysis

Fluorescence image analysis in biochemical science often involves the complex tasks of identifying samples for analysis and calculating the desired information from the intensity traces. Analyzing giant unilamellar vesicles is one of these tasks. Researchers need to identify many vesicles to statistically analyze the degree of molecular interaction or state of molecular organization on the membranes. This analysis is complicated, requiring a careful manual examination by researchers, so automating the analysis can significantly aid in improving its efficiency and reliability. We developed an intelligent analysis routine based on the 3D information of whole z-stack images. The program identifies the valid vesicles to analyze and calculates the desired data automatically. The program can examine the amount of protein binding on the membranes and determine the state of domain phase separation by calculating the fluorescence intensity trace along the membranes. We also show that the method can easily be applied to similar analyses, such as intensity analysis of phase-separated protein droplets. A deep learning-based classification approach enables the identification of vesicles even from relatively complex samples. We demonstrate that the proposed artificial intelligence-assisted classification can further enhance the accuracy of the analysis close to the performance of manual examination.

sweetD: An R package using Hoeffding’s D statistic to visualise the dependence between M and A for large numbers of gene expression samples

SummaryMA plots are frequently used to examine the relationship between gene abundance and differences in gene expression between two samples. In good quality samples without batch effects or outliers, there is generally no or little relationship between intensity and log fold difference. As the number of MA plots increases quadratically with the number of samples, for large datasets the number of potential MA plots becomes prohibitively high for manual examination. Here we present an R package calculating and visualising the dependence between abundance and log fold difference for large numbers of samples using Hoeffding’s D statistic.Availability and implementationsweetD is currently available as an R package on Github https://github.com/amberjoybarton/sweetD.

Informative application of the instrumental diagnosis of scar integrity in the experimental model of the uterine cavity

Currently, the frequency of deliveries by cesarean section is steadily increasing. Independent childbirth in women with a uterine scar is one of the mechanisms of its reduction. Meanwhile, the delivery of this category of patients through the natural birth canal requires the development of a safe and effective method for assessing scar integrity in the early postpartum period. For this purpose, a new instrumental diagnostics method has been developed, with its information content compared to manual examination data in an experimental model of the uterus. Each model was made of a bovine heart and had three defects with diameters less than 0.5 cm, 0.5-0.8 cm and 0.8-1.2 cm. Defects with diameters less than 0.5 cm were not detected by any of the methods studied. Defects with a diameter of 0.5-0.8 cm were detected using the developed device in 90 % (45/50) of cases and using manual examination in 44 % (22/50) of cases (2 = 23.93, p 0.001), defects with a diameter of 0.8-1.2 cm being detected in 100% (50/50) and 98% (49/50) of cases, respectively (2 = 1.01, p = 0.32). The information content of the instrumental model exceeds that of the manual study by 1.34 times (RR = 1.34, 95 % CI 1.09-1.65, p 0.05). Therefore, the possibility of testing this device in clinical trials needs to be considered.