DEVELOPMENT AND OPTIMIZATION OF AI-BASED BIOMEDICAL IMAGING SYSTEM FOR EARLY DISEASE DETECTION AND DIAGNOSIS

The present study focused on the development and optimization of an artificial intelligence (AI)-based biomedical imaging system for early disease detection and diagnosis. A quantitative, experimental, and computational research design was adopted to develop an integrated framework for automated biomedical image processing and classification. The research methodology involved the collection and analysis of biomedical images representing normal and abnormal diagnostic categories. A total of 1,000 images, comprising 600 normal and 400 abnormal images, were included in the analytical framework. Image preprocessing, resizing, normalization, noise reduction, contrast enhancement, and feature extraction were performed using R software and appropriate image-processing packages. Machine learning algorithms, including Logistic Regression, Decision Tree, Random Forest, Support Vector Machine, and Convolutional Neural Network, were developed and comparatively evaluated. Model optimization and validation were conducted through cross-validation and diagnostic performance assessment. R packages, including EBImage, caret, randomForest, e1071, pROC, and ggplot2, supported image processing, model development, statistical analysis, and graphical visualization. Data analysis focused on diagnostic accuracy, sensitivity, specificity, precision, recall, F1-score, AUC-ROC, and computational efficiency. The findings demonstrated the potential of optimized deep learning models to improve biomedical image classification and diagnostic performance. The significance of the study lies in establishing an integrated and reproducible R-based computational framework that may assist healthcare professionals in identifying abnormal imaging patterns, strengthening diagnostic support, and facilitating early disease detection.