Leg bone fractures require accurate and timely diagnosis to determine appropriate treatment. This research proposes an Artificial Intelligence (AI) based deep learning framework for automated leg bone fracture detection and severity assessment using X-ray images. Initially, a Median Filter is applied as a preprocessing technique to remove noise and enhance fracture-related structures. The preprocessed images are then analyzed using Convolutional Neural Network (CNN) and Artificial Neural Network (ANN) algorithms for feature extraction, fracture detection, and classification. The proposed system identifies fractured and non-fractured bones and assesses the severity or stage of the fracture. Based on the detected fracture characteristics, the system provides computeraided treatment recommendations, including conservative management, supportive medication, or surgical intervention when clinically appropriate. The proposed framework aims to improve fracture detection accuracy, reduce diagnostic time, and support healthcare professionals in making effective treatment decisions.

Objectives: This study examines the impact of chemical fertilizers and pesticides used in the cultivation of vegetables, ground nuts, and paddy on human health, soil quality, and the environment. Specifically, it examines the risk of organ damage, including kidney failure, liver impairment, vision problems, heart and lung complications, and the potential development of cancer. The study also evaluates the benefits of natural and organic farming practices, such as the use of animal dung and garlic-based bio-fertilizers, in mitigating these risks and promoting sustainable agriculture. Methods: A comprehensive analysis was conducted using advanced deep learning techniques. Convolutional Neural Networks (CNN) were employed to classify agricultural data and detect contamination patterns in crops caused by chemical fertilizers and pesticide residues. Long Short-Term Memory (LSTM) networks were used to model temporal trends, capturing the cumulative effects of chemical usage on soil health and long-term human exposure. Random Forest classifiers were integrated to identify and prioritize critical features that contribute to organ-related health risks and environmental degradation. A comparative analysis was conducted to evaluate the effectiveness of natural farming practices in mitigating these risks. Findings: The results indicate that crops cultivated with chemical inputs significantly threaten human health. Regular consumption can lead to kidney failure, liver damage, impaired vision, cardiovascular and respiratory complications, and increased cancer susceptibility. Furthermore, the use of chemicals reduces soil fertility, disrupts microbial ecosystems, and negatively impacts the environmental cycle. Conversely, organic farming practices, including the use of animal dung and garlic-based bioproducts, improved soil quality, preserved ecological balance, and produced safer food, thereby protecting human organs. The study demonstrates that combining deep learning-based predictive risk analysis with sustainable farming methods provides an effective strategy to minimize health hazards, maintain soil fertility, and ensure environmental sustainability.

Background: Chronic kidney disease often develops without any symptoms, and the awareness in the population has not become so far despite advances in risk stratification and kidney-protective therapy. It is uncertain whether the contemporary U.S. population recognizes albuminuric and reduced-filtration phenotypes differently. Methods: We examined data from the National Health and Nutrition Examination Survey (NHANES) conducted between August 2021 and August 2023 in adults aged ≥20 years with serum creatinine, urinary albumin to creatinine ratio (UACR), and kidney awareness questionnaire data. The 2021 CKD-EPI creatinine equation was used to estimate the eGFR. CKDcompatible abnormality in the laboratory was eGFR <60 mL/min/1.73 m2 and/or UACR ≥30 mg/g at the single NHANES examination. In this case, awareness was defined as personally reporting having been told by a health professional that the participant has weak or failing kidneys. The weight for phlebotomy, cycle-specific, strata and primary sampling units determined survey analyses. Results: Among 5298 adults, there were 918 laboratory abnormality; weighted prevalence was 14.0 % (95% CI, 12.5%15.6%). Of the adults who had the disease, only 12.9% knew about it (95% CI, 9.9-16.0%), meaning 87.1% remained unrecognized. In isolated albuminuria, awareness was 3.1%, in isolated eGFR <60 it was 17.9%, while in both was 39.8%. In totally modified models, compared with isolated albuminuria, awareness was higher with isolated reduced eGFR (OR 7.42, 95% CI 2.85–19.30) and the merged phenotype (OR 27.63, 95% CI 11.19–68.21). Being in the A3 albuminuria category or having severe filtration impairment (eGFR <45) was independently associated with a higher awareness; however, even in the most severe combination, awareness was under 50%. Conclusion: In contemporary US adults, kidney dysfunction remains mostly unrecognized, especially when the abnormality is albuminuria with preserved eGFR. Tests of serum creatinine will not close the gap, which is substantial, although dual assessment with eGFR and UACR needs to be more widespread.

Background: There is an increasing research on automated recognition and assessment of vitiligo using AI. Nevertheless, the high classification performance may be partly due to technical differences between classes rather than disease-related morphology. Objective: The aim of this study was to evaluate the integrity as well as the technical characteristics of a publicly available vitiligo image dataset. Besides that, it aimed to determine whether non-clinical image metadata alone can discriminate vitiligo from healthy-image labels. Methods: We performed a systematic data-quality audit of a publicly available dataset consisting of 3,628 images classified as vitiligo or healthy. We checked the font quality, total number of classes in the dataset, file format, width and height dimensions, file size, aspect ratio, pixel count, and duplication. MD5 hashing was used to identify duplicate files and remove them prior to the modeling stage. Mann-Whitney U test and rank-biserial correlations were used to compare variables and groups on technical characteristics. A logistic regression was calculated on image width, height, file size, aspect ratio, megapixel count (mp), and bytes per pixel, as a negative-control experiment. Pixels in the image as well as the dermatological features are not included. We evaluated performance by stratified five-fold cross-validation, and estimated 95% confidence intervals (CIs) using 2,000 bootstraps. Results: All images (3,628) were legible including the (2,100) vitiligo and (1,528) healthy images. A total of forty redundant exact copies were identified, all belonging to the class vitiligo, leaving only 3588 unique images. Technical characteristics showed major differences between classes. In the healthy category, images had an average resolution of 1600 × 1200 pixels, whereas in the vitiligo category, the average resolution was 416 × 416 pixels. The median file size was 66,374 and 17,323 bytes respectively. The classes significantly differed with regard to all the technical variables examined (p < .001). Interestingly, the six-variable metadata-only model achieved an area under the curve (AUC) of 0.942, with an accuracy of 0.875, sensitivity of 0.917, and specificity of 0.817 (95% CI: 0.933-0.950). A model solely based on image width and height achieved an AUC of 0.929 Conclusion: Technical attributes of the images in the public vitiligo dataset are not clinically-related. As a result, high internal classification correctness achieved by the data set may partly relate to source or acquisition shortcuts rather than vitiligo-specific morphology. Before determining clinical claims based on models developed using such data, we should audit the dataset for issues like duplication or deviation from established sources and standards.

Portal vein thrombosis (PVT) is an uncommon vascular disorder characterized by partial or complete obstruction of the portal vein by a thrombus. It may result in portal hypertension, splenomegaly, and, in chronic cases, secondary liver cirrhosis. Oral contraceptive pills (OCPs) are widely used hormonal agents and are known to increase the risk of thromboembolic events due to their prothrombotic effects. However, portal vein thrombosis associated with oral contraceptive use progressing to cirrhosis is rarely reported.
We present the case of a 26 year old woman who presented to the emergency department with complaints of abdominal pain and jaundice. Laboratory evaluation revealed anemia, thrombocytopenia, mild coagulopathy, and abnormal liver function tests. Imaging studies demonstrated heterogeneous liver parenchyma, lobulated hepatic contours, and splenomegaly consistent with chronic liver disease and portal hypertension. Detailed medical history revealed prolonged oral contraceptive use after her second pregnancy, which was considered a contributing factor to the development of portal vein thrombosis and secondary liver cirrhosis. The patient was admitted to the gastroenterology department for further evaluation and management.
This case highlights the importance of considering portal vein thrombosis as a potential complication of oral contraceptive use and emphasizes the need for careful evaluation of thrombotic risk factors in young women presenting with liver disease.