BRAIN DISEASE PREDICTION

This research proposes a comprehensive machine learning framework for brain age estimation and neurodegenerative disease classification to support early diagnosis of Parkinson's disease and Alzheimer's disease.The framework integrates regression-based brain age prediction with deep learning-based disease classification using structural MRI data.In the initial phase, multiple regression models-including Linear Regression, Ridge Regression, Lasso Regression, Random Forest Regression, and Support Vector Regressionare evaluated for estimating brain age.Performance is assessed using Mean Absolute Error (MAE) and R² metrics, with Ridge and Lasso Regression achieving the most accurate predictions.Deviations between predicted brain age and chronological age are identified as potential early indicators of neurodegenerative conditions.Based on these findings, a convolutional neural network (CNN) is employed to classify brain images into healthy aging, Parkinson's disease, and Alzheimer's disease.The CNN model demonstrates strong and consistent performance across accuracy, precision, recall, and F1-score metrics, achieving an overall classification accuracy of 99.69%.Robust preprocessing, feature extraction, and model optimization ensure clinically relevant and reliable results.The proposed framework highlights the effectiveness of combining brain age estimation with deep learning for early disease detection and clinical decision support.Future research may focus on improving model generalization, incorporating interpretability techniques, and extending the framework to include longitudinal and multimodal data for enhanced understanding of disease progression.