Detection of Parkinson disease for handwriting dataset using deep learning algorithms
A. Malathi, Ramalakshmi Ramar, Vaibhav Gandhi · 2023
The most common neurodegenerative diseases are Parkinson's disease. People who have reached the age of 70 or older are affected. Parkinson's disease is the second incurable illness after Almeria's disease. The numerous characteristics that distinguish between people with Parkinson's disease and those who are healthy, and earlier diagnosis of the disease results in reduced disease severity. These are the typical signs of Parkinson's disease, which include both motor and non-motor deficiency in patients. The handwritten data set, which includes spiral and wave drawings, was retrieved from UCI equipment. EfficientnetB3 provides high accuracy in comparison to the current CNN model. To increase efficiency, EfflcientnetB3 shrinks the parameter size and equally scales each dimension using a predetermined set of scaling dimensions. The proposed model Adam –EfficientNet achieves 98.3 % which is the best accuracy compared to all other existing CNN models.