A Hybrid Approach Based on CNN and Extremely Randomized Trees for the Diagnosis of Alzheimer's Disease Stages from MRI Images

Alzheimer's disease, a widespread form of dementia, impairs the nervous system and leads to the decline of various brain functions, including memory loss. Early detection of Alzheimer's disease is an area of significant research interest, as early diagnosis plays a crucial role in improving patient care and treatment outcomes. In this paper, we propose an approach that combines a pre-trained Convolutional Neural Network model, namely Inception-V3for feature extraction with a supervised machine learning technique, specifically Extremely Randomized Trees, is employed for classification purposes. This classifier enables the differentiation of various stages of Alzheimer's disease. The obtained results are promising and demonstrate the effectiveness of our approach when evaluated with two benchmark datasets. The accuracy is estimated to 92.31% and 98.57% with ADNI and MIRIAD dataset, respectively.