METHODS: In this work, we developed deep convolutional neural network (CNN) based heterogeneous ensemble models for automated analysis of renal histopathological images without detailed annotations. The proposed method would first segment the histopathological tissue into patches with different magnification factors, then classify the generated patches into normal and tumor tissues using the pre-trained CNNs and lastly perform the deep ensemble learning to determine the final classification. The heterogeneous ensemble models consisted of CNN models from five deep learning architectures, namely VGG, ResNet, DenseNet, MobileNet, and EfficientNet. These CNN models were fine-tuned and used as base learners, they exhibited different performances and had great diversity in histopathological image analysis. The CNN models with superior classification accuracy (Acc) were then selected to undergo ensemble learning for the final classification. The performance of the investigated ensemble approaches was evaluated against the state-of-the-art literature.
RESULTS: The performance evaluation demonstrated the superiority of the proposed best performing ensembled model: five-CNN based weighted averaging model, with an Acc (99%), specificity (Sp) (98%), F1-score (F1) (99%) and area under the receiver operating characteristic (ROC) curve (98%) but slightly inferior recall (Re) (99%) compared to the literature.
CONCLUSIONS: The outstanding robustness of the developed ensemble model with a superiorly high-performance scores in the evaluated metrics suggested its reliability as a diagnosis system for assisting the pathologists in analyzing the renal histopathological tissues. It is expected that the proposed ensemble deep CNN models can greatly improve the early detection of renal cancer by making the diagnosis process more efficient, and less misdetection and misdiagnosis; subsequently, leading to higher patients' survival rate.
METHODS: A systematic review and meta-analysis of studies that reported the global prevalence of gallstones in pregnancy was conducted. PubMed, Scopus, Web of Science, Embase, ScienceDirect, and Google Scholar were searched for studies published up to September 2022.
RESULTS: In a review of 31 studies with a sample size of 190,714 people, the I2 heterogeneity test showed high heterogeneity (I2 = 98.8%). Therefore, the random effects method was used to analyze the results. The prevalence of gallstones was reported as 3.6% (95% CI: 1.9-6.7%). The highest prevalence of gallstones by continent was reported in America, at 6.8% (95% CI: 4.2-10.8%). The Egger test showed no evidence of publication bias (p = 0.609).
CONCLUSION: Based on the results of this study, health policymakers should emphasize to the target community and the medical staff dealing with pregnant women the importance of screening for gallstones during pregnancy.
OBJECTIVE: This narrative review's objective was to explore the use of AOM in relation to their medical indications, efficacy, and cardiovascular safety.
METHODS AND MATERIALS: We have conducted a narrative review of the literature on approved/non-approved AOM used for obesity and overweight. We have shed light on the emerging trials of therapies and evolving remedies.
RESULTS: Recently, there has been an enormous change in the use of AOM with high consumption that deserves extensive surveillance for the long-term consequences and impact on social, mental, and physical health. Nearly six AOMs and combined therapy are approved by the Food and Drug Administration. The recent guidelines for obesity management have shifted the focus from weight loss to goals that the patient considers essential and toward targeting the root cause of obesity.
CONCLUSION: The use of AOM increased enormously despite its sometimes-dubious safety and ineffectiveness. The public and medical professionals should be vigilant to the real-world benefits of anti-obesity drugs and their achieved effectiveness with an improved safety profile.
METHODS: The data were obtained from the National Health and Morbidity Survey (NHMS) which was conducted from September to October 2020. A cross-sectional survey with five structured questionnaires using the method of computer-assisted telephone interviews (CATI) was used to collect data. The socio-demographic characteristics such as age, gender, ethnicity, nationality, marital status, educational level, and occupation were recorded. Data were analysed using STATA SE Version 16. Associations between variables were tested using chi-square and logistic regression, with the level of statistical significance set at p
OBJECTIVE: To evaluate differences in RFs estimation based on unenhanced computerized tomography (CT) versus X-rays/ultrasound after retrograde intrarenal surgery (RIRS) for kidney stones.
DESIGN: A retrospective analysis of data from 20 centers of adult patients who had RIRS was done (January 2018-August 2021).
METHODS: Exclusion criteria: ureteric stones, anomalous kidneys, bilateral renal stones. Patients were divided into two groups (group 1: CT; group 2: plain X-rays or combination of X-rays/ultrasound within 3 months after RIRS). Clinically significant RFs (CSRFs) were considered RFs ⩾ 4 mm. One-to-one propensity score matching for age, gender, and stone characteristics was performed. Multivariable logistic regression analysis was performed to evaluate independent predictors of CSRFs.
RESULTS: A total of 5395 patients were included (1748 in group 1; 3647 in group 2). After matching, 608 patients from each group with comparable baseline and stone characteristics were included. CSRFs were diagnosed in 1132 patients in the overall cohort (21.0%). Post-operative CT reported a significantly higher number of patients with RFs ⩾ 4 mm, before (35.7% versus 13.9%, p
METHODS: This was a retrospective cross-sectional study conducted on specialist medical reports written from 2009 to 2019, involving patients who survived after TBI from RTA. The functional outcome was assessed using the Glasgow Outcome Scale-Extended (GOSE). Factors associated with good outcome were analysed via logistic regression analysis. Multivariate logistic regression analysis was used to derive the best fitting Prediction Model and split-sample cross-validation was performed to develop a prediction model.
RESULTS: A total of 1939 reports were evaluated. The mean age of the study participants was 32.4 ± 13.7 years. Most patients were male, less than 40, and with average post RTA of two years. Good outcome (GOSE score 7 & 8) was reported in 30.3% of the patients. Factors significantly affecting functional outcome include age, gender, ethnicity, marital status, education level, severity of brain injury, neurosurgical intervention, ICU admission, presence of inpatient complications, cognitive impairment, post-traumatic headache, post traumatic seizures, presence of significant behavioural issue; and residence post discharge (p<0.05). After adjusting for confounding factors, prediction model identified age less than 40, mild TBI, absence of post traumatic seizure, absence of behaviour issue, absence of cognitive impairment and independent living post TBI as significant predictors of good functional outcome post trauma. Discrimination of the model was acceptable (C-statistic, 0.67; p<0.001, 95% CI: 0.62-0.73).
CONCLUSION: Good functional outcome following TBI due to RTA in this study population is comparable to other low to middle income countries but lower than high income countries. Factors influencing outcome such as seizure, cognitive and behavioural issues, and independent living post injury should be addressed early to achieve favourable long-term outcomes.