Detecting Parkinson's Disease with Image Classification
S. Kanagaraj, M. Hema, M. Nageswara Guptha, V. Namitha
Abstract
The non-curable neurological disorder that affects the motor system is known as Parkinson disease. When Parkinson disease is detected earlier, then it can diagnose, and we can get a quick relief but not permanent. The neurons segregate a chemical called dopamine. That helps for transmitting the signs to the other neurons in the brain. When the dopamine flow starts to fall, then the PD occurs. This makes the patients to, resting tremors, bradykinesia and rigidity problems. Here machine-learning dramatizations position in patterns tag in biomedical sciences. The PD mainly attack the motor system so that can be analysed by the Magnetic Resonance Imaging (MRI) scan, one can detect and predict the disease. In this paper, with MRI scan the Parkinson's disease is detected by using CNN, VGG-16 model and ResNET-50. The VGG-16 and ResNet-50 are compared and find the best model based on the accuracy.