Volume 14 | Issue 4 |

 Abstract

Among the leading contributors to global cancer mortality, lung cancer stands out due to the difficulty of de tecting it before it reaches an advanced stage. Manual review of Computed Tomography (CT) scans demands considerable radiologist effort and remains susceptible to perceptual errors. This work introduces an end-to-end deep learning framework that automates the classification of lung CT images into benign and malignant categories. The system builds on MobileNetV2 --acompact yet powerful convolutional architecture pretrained on ImageNet -- and adapts it to the medical imaging do main through a structured two-phase fine-tuning procedure. A dataset of 495 CT images was used for training, validation, and testing, with class imbalance managed via adaptive sample weighting. Comprehensive image preprocessing and controlled augmentation were applied to improve model generalization. The resulting model attained 99.38% accuracy, 99.08% precision, 100.00% sensitivity, 98.15% specificity, a 99.54% F1-score, and a ROC-AUC of 0.9935 on 162 held-out test images. A browser accessible Streamlit application was built around the trained model, enabling clinicians to obtain real-time predictions directly from uploaded scans without any machine learning background. The findings confirm that combining transfer learning with carefully designed class-balancing strategies yields substantially better outcomes than naive from-scratch CNN training.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Lung Cancer Detection, Convolutional Neural Networks, MobileNetV2, Transfer Learning, CT Scan Analysis, Medical Imaging, Clinical Decision Support, Web Deployment, Deep Learning

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Abstract: This review paper presents a comprehensive analysis of bridge piers with varying geometric shapes, focusing on their structural performance under standard loading conditions for a 12.3 m span as per RDSO-B-11775 guidelines. The primary aim is to compare different pier configurations and identify the most efficient shape based on structural behavior. Key parameters such as load-bearing capacity, displacement, bending moments, and stress distribution are critically reviewed from existing studies. Different geometric shapes, including rectangular, circular, and hammerhead piers, are evaluated considering dead load, live load, and impact load conditions. The review highlights how geometry significantly influences stiffness, stress concentration, and overall structural response. Circular piers are found to provide uniform stress distribution and improved performance, whereas rectangular piers exhibit higher stiffness but suffer from stress concentration at corners. Hammerhead piers offer advantages in load transfer efficiency for superstructures. The findings from various studies indicate that selecting an optimal pier geometry is essential for enhancing stability, durability, and safety. This review contributes to better design understanding by comparing different pier shapes and identifying circular piers as a more effective and reliable option for bridge construction.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Keywords: Bridge Piers, Geometric Shapes, Structural Performance, STAAD Pro, Cyclic Loading, RDSO-B-11775 etc.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Keywords: Prevalence, Diabetes mellitus, Rural and urban population

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Ischemicstrokesubtypingwasnotonlyhighlyvaluableforeffec tiveinterventionandtreatment, but also important to the prognosis of ischemic stroke. The manual adjudication of disease classification was time-consuming, error-prone, and limits scaling to large datasets. In this study, an integrated machine learning approach was used to classify the subtype of ischemic stroke on The International Stroke Trial (IST) dataset. We considered the common problems of feature selection and prediction in medical datasets. Firstly, the importances of features were ranked by the Shapiro-Wilk algorithm and Pearson correlations between features were analyzed. Then, we used Recursive Feature Elimination with Cross-Validation (RFECV), which incorporated linear SVC, Random-Forest-Classifier, Extra-Trees-Classifier, AdaBoost-Classifier.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

We considered the common problems of feature selection and prediction in medical datasets.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Vaginal infections are among the most common health issues affecting women of reproductive age, primarily caused by an imbalance in the normal vaginal microflora. Conditions such as bacterial vaginosis, candidiasis, and trichomoniasis are often associated with symptoms like itching, irritation, inflammation, and abnormal discharge. Conventional treatments, including antibiotics and antifungal agents, though effective, are frequently associated with side effects, recurrence, and the development of drug resistance. Therefore, there is a growing need for safer, more effective, and natural therapeutic alternatives. The present study focuses on the formulation and evaluation of a herbal vaginal gel incorporating medicinal plant extracts known for their antimicrobial, anti-inflammatory, and soothing properties. Selected herbal ingredients such as Aloe vera, Azadirachta indica (Neem), Ocimum sanctum (Tulsi), Curcuma longa (Turmeric), and pomegranate peel extract were utilized due to their well-documented pharmacological activities. The extracts were prepared using suitable extraction techniques and incorporated into a gel base formulated with appropriate gelling agents to ensure optimal viscosity, spreadability, and bioadhesion. The formulated gels were evaluated for various physicochemical parameters including pH, viscosity, homogeneity, spreadability, and extrudability to ensure suitability for vaginal application. Microbiological studies were conducted to assess antimicrobial activity against common vaginal pathogens such as Candida albicans and Escherichia coli. Stability studies were also performed under different storage conditions to determine the shelf life and consistency of the formulation. The results indicated that the herbal vaginal gel exhibited satisfactory physicochemical characteristics and significant antimicrobial activity, suggesting its potential effectiveness in managing vaginal infections. The formulation also demonstrated good stability and compatibility of herbal constituents within the gel base. The use of herbal ingredients provides additional advantages such as reduced side effects, improved patient compliance, and enhanced therapeutic efficacy.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Bacterial vaginosis, Punica Granatum, Antimicrobial and antifungal activity, Antimicrobial study

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Buddhist medicine, monastic hospitals, faith-based pandemic response, integrative medicine, public health, Vietnam Buddhist Sangha, narrative scoping review

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

?? ??????? ????????? ????????? ?? ????????? ?? ????? ?????????, ?????????? ?? ????????????? ???????? ?? ?????? ?? ???? ???????? ????? ?? ???????? ?????? ?? ???????? ???? ??? ?????? ?? ????? ???????? ?? ?????? ???? ?? ?? ????????? ?? ????? ??????? ?????? ??? ???????, ??????? ??????, ???????-???? ?????, ????????? ????? ??? ?????????? ??????? ???? ?????? ?? ??? ?????? ???????? ????? ?????? ????? ???? ?????????--??????? ???????????, ?? ?? ???? ?? ??? ???--??? ???????????, ??????? ????, ??????? ?????? ?? ??????? ???????? ?? ??????????? ?????? ???? ??? ??? ????????? ??? ???, ?? ??????? ?? ???? ??????? ?????????? ? ????? ?? ????? ?????????? ???????? ?? ?????? ???? ???? ??? ??? ?? ?????????? ???? ??? ?? ?? ????????? ?? ????????? ???? ?? ????????? ????? ?? ?????????? ?? ???????? ???? ??? ?? ????? ???? ?? ???? ??????? ?? ???????? ????? ???? ??????????? ?? ??????? ??????????? ??? ????? ???? ??? ??? ??, ???? ??????? ??? ??????? ?? ?????? ?? ?????????? ?? ??????? ?????? ?????????? ?????? ?? ?? ?????? ??? ?? ?????? ?? ???????? ?? ??????? ?? ?? ????????? ?? ??????? ? ???? ????? ??????? ?? ???????? ???? ?????? ???? ??, ????? ?? ???????? ???, ?????????? ?????? ?? ??????? ????? ?????? ???????? ?? ??? ?? ?????????? ????? ??, ????? ???? ?????? ?? ?? ????????? ?? ????? ??????? ?? ??? ??? ?? ???? ??? ?? ????????? ????? ???? ????

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

????? ????: ????????? ?????????, ???????, ??????? ????, ??? ????????, ??????? ????, ???????????, ?? ?? ????, ???????, ??? ???, ?????????? ???????, ?????? ??????, ?????? ?????????, ??????? ??????, ?????????? ?????, ?????? ????????

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This project presents an Automated Fake News Identification System using machine learning techniques to detect misleading information in digital content. It addresses the rapid spread of fake news on social media and online platforms. The system applies Natural Language Processing for text preprocessing, including tokenization and stopword removal. TF-IDF feature extraction converts text into numerical form for classification. A Flask-based interface allows users to input news and receive instant predictions. The model is evaluated using accuracy, precision, and recall to ensure reliable and scalable misinformation detection.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Fake News Detection, Machine Learning, Natural Language Processing, Text Preprocessing, TF-IDF, Classification, Social Media, Flask Interface, Accuracy Precision Recall, Automated Fact-Checking

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This study looks at how college students use educational mobile apps in their daily academic lives. The research is based on primary data from 100 students who answered a structured questionnaire. It focuses on factors like age, gender, academic level, time spent on apps, types of apps used, and students' opinions. The data was analysed using basic business analytics tools like percentages and simple tables. Most students prefer to use smartphones for educational apps. Video-based learning platforms are the most popular. Many students also say these apps help them understand their subjects better and improve their academic performance. The study also tests three ideas about how often students use apps, how much time they spend, and what types of apps they use. The results show these factors are linked to student engagement and academic level. These findings can help educators, app developers, and institutions improve digital learning systems.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Mobile apps, educational apps, college students, usage patterns, business analytics.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Education being the backbone of any society has always been the main subject of that society's rulers, philosophers, and intellectuals. From time to time in every society, there have been arguments, debates, and discussions on how education should be conducted so that the objectives of that society can be achieved. The final task of formulating these arguments, debates, and discussions in the form of policies and implementing them among their students is done by the teachers, who also determine the form of education of that society. With time, along with traditional/conventional mediums, with the help of technology, a large section is being educated according to their social, and professional needs through distance or online mediums. All types of academic education of the society have been included in the National Education Policy 2020, in this, efforts have been made to fulfil the objectives related to education while using the methods of distance education / online mediums. In the NEP 2020, policy efforts and suggestions have also been made regarding professional competence in teacher education for teachers of traditional/ conventional medium, but there seems to be a lack of such efforts in NEP 2020 for teachers and employees working in distance education. Even big education and teacher education platforms like DIKSHA (Digital Infrastructure for Knowledge SHAring) and SWAYAM (Study Webs of Active- learning for Young Aspiring Minds) have not been able to provide any courses in this direction. In this research paper, an attempt has been made to clarify what is there for teacher education in this policy and what kind of provisions have been made for the needs of teacher education, especially for distance education and to suggest solutions from our side.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

National Education Policy 2020, Distance Education, DIKSHA (Digital Initiative for Knowledge SHAring), CPD (Continuous Professional Development), MOOC (Massive Open Online Courses), CRC (Cluster Resource Centre)

  License

Creative Commons Attribution 4.0 and The Open Definition