Omdena / ML engineer
Analyzing Brain Scan Images for the Early Detection and Diagnosis of Alzheimer's Disease
• Gathered MRI brain scans from Kaggle and the ADNI database. Preprocessed the ADNI dataset through techniques such as resampling, noise reduction filtering, and intensity normalization. Augmented the data to enhance model accuracy. • Addressed class imbalance using Synthetic Minority Over-sampling Technique (SMOTE). Established a custom Convolutional Neural Network (CNN) architecture, employing KerasTuner for hyperparameter tuning. Achieved an impressive model accuracy of 99.22%.