A Transfer Learning Approach for Neurodegenerative Disease Classification from Brain MRI Images: Distinguishing Alzheimer's, Parkinson's, and Control Cases

Like many other critical medical conditions, different neurogenerative diseases, including Alzheimer's and Parkinson's diseases, need to get diagnosed in the primary stage. Deep learning algorithms show excellent performance in detecting neurodegenerative diseases from images obtained by magnetic resonance imaging (MRI) of the human brain. Recently, transfer learning has also shown promising outcomes in the classification of neurological conditions using brain MRI data. Here, we examine the efficacy of the transfer learning strategy utilizing four distinct CNN architectures, namely EfficientNetB0, ResNet50, InceptionV3,and Xception. Used dataset for the study has three classes; Parkinson's disease (PD), Alzheimer's disease (AD), and control (healthy). The study compares the accuracy, precision, recall, and F1- score metrics for the investigated CNN models. The result demonstrates that the EfficientNetB0 model shows the best training and testing accuracy, reaching an accuracy as high as 99.4%.