International Journal of Artificial Intelligence in Medical Issues

Obesity Prediction with Machine Learning Models Comparing Various Algorithm Performances

Yudha Islami Sulistya — 2025-05-30

Obesity poses a significant global health risk due to its links to conditions such as diabetes, cardiovascular disease, and various cancers, underscoring the need for early prediction to enable timely intervention. This study evaluated the performance of seven machine learning algorithms—Logistic Regression, Decision Tree, Random Forest, ExtraTrees, Gradient Boosting, AdaBoost, and XGBoost—in predicting obesity using health and lifestyle data. The models were assessed based on accuracy, precision, recall, and F1-score, with hyperparameter tuning applied for optimization. The results confirmed that the ExtraTrees Classifier was the best performer, achieving an accuracy of 92.6%, precision of 92.7%, recall of 92.8%, and F1-score of 92.7%. Both Random Forest (91.3% accuracy) and XGBoost (89.9% accuracy) also exhibited strong predictive abilities. In contrast, models like Logistic Regression (74.3% accuracy) and AdaBoost (73.0% accuracy) showed lower effectiveness, emphasizing the advantages of ensemble methods such as ExtraTrees in delivering accurate obesity predictions. These findings suggest that ensemble models provide a promising approach for early diagnosis and targeted healthcare interventions.

Classification of Multi-Region Bone Fractures from X-ray Images Using Transfer Learning with ResNet18

Rasni Alex — 2025-05-30

Fracture detection in radiographic images is a critical task in orthopaedic diagnostics, often requiring timely and accurate interpretation by medical professionals. However, manual evaluation of X-rays is time-consuming and prone to subjective bias. This study proposes an automated deep learning approach for binary classification of bone fractures using a pre-trained ResNet18 architecture. The model was trained and validated on a multi-region X-ray dataset consisting of 10,580 images categorized into fractured and non-fractured classes. To improve generalization, data augmentation techniques such as rotation and horizontal flipping were applied during pre-processing. The final model achieved a validation accuracy of 97.59%, with high true positive and true negative rates as confirmed by the confusion matrix analysis. The results demonstrate the effectiveness of transfer learning in handling radiographic image classification tasks while maintaining computational efficiency. This research contributes to the development of reliable and scalable computer-aided diagnostic tools that can support clinical decision-making, especially in environments with limited resources.

Effect of Screen Time on Glaucoma

Mahule — 2025-05-30

Glaucoma, characterized by elevated intraocular pressure (IOP) and optic nerve damage, encompasses various types with distinct pathogenic mechanisms. Research has identified key factors influencing glaucoma, such as environmental influences, stress, and age-related factors. This study focuses on the impact of stress on IOP levels in glaucoma patients and evaluates different machine learning (ML) models for enhanced glaucoma detection using OCT and Color Fundus images. Additionally, I explore the environmental implications of elevated IOP, emphasizing lifestyle interventions like yoga to potentially reduce IOP levels. As a practical application, I propose the development of a dedicated mobile app as a digital wellness program for glaucoma patients.

Predicting Thyroid Cancer Recurrence After Radioactive Iodine Therapy Using Random Forest and Neural Network Models

Rosmasari — 2025-05-30

Thyroid cancer recurrence following Radioactive Iodine (RAI) therapy remains a clinical concern, necessitating accurate and timely risk prediction to guide post-treatment management. This study aims to evaluate the effectiveness of machine learning models—Random Forest and Neural Networks—in predicting recurrence using a structured clinical dataset consisting of 383 patient records and 13 diagnostic and pathological attributes. All categorical features were encoded ordinally, and the dataset was partitioned into training and testing sets with appropriate normalization for neural network processing. Both models were evaluated using standard metrics including accuracy, precision, recall, and F1-score. The Random Forest model achieved an accuracy of 97.39%, outperforming the Neural Network which recorded 93.04%. Moreover, Random Forest showed better recall in detecting recurrence cases, making it a more suitable model for clinical application. These results demonstrate that machine learning, particularly ensemble-based methods, can offer a practical and interpretable solution for recurrence prediction, supporting data-driven decision-making in thyroid cancer follow-up care.

Automated Diagnosis of Benign Prostatic Hyperplasia Using Deep Learning on RGB Prostate Images

Lukman Syafie — 2025-05-30

Benign Prostatic Hyperplasia (BPH) is a prevalent non-cancerous enlargement of the prostate gland in aging men, often requiring early diagnosis to prevent urinary complications and improve patient outcomes. Traditional diagnostic procedures are limited by subjectivity and accessibility, especially in under-resourced regions. This study proposes an automated diagnostic approach using a deep learning model based on DenseNet121 to classify RGB prostate images into BPH and normal categories. A region-specific dataset consisting of 176 labeled RGB images, collected from a clinical facility in Bangladesh, was used to train and evaluate the model. Pre-processing included image resizing, normalization, and data augmentation to enhance generalization. Transfer learning was employed to fine-tune the model, which was trained over 10 epochs using the Adam optimizer and cross-entropy loss. The model achieved a best validation accuracy of 94.12%, with a recall of 72.2% for BPH detection, demonstrating its ability to identify pathological patterns in simple imaging modalities. Despite challenges such as dataset size and imbalance, the findings indicate that RGB image-based deep learning models can support clinical diagnosis of BPH in low-resource settings. This work contributes a lightweight, accessible solution for prostate disease screening and provides a foundation for future research on scalable AI-assisted diagnostics.