PARKINSON’S DISEASE ANALYSIS USING DEEP LEARNING AND VGG-16 MODEL-BASED APPROACHES

Abstract

Parkinson's disease (PD) is an intricate neurodegenerative condition that impacts areas like the Substantia Nigra (SN), Red Nucleus (RN), and locus coeruleus (LC). Analyzing MRI data from PD patients necessitates anatomical structural landmarks for spatial standardization and structural partitioning. Machine Learning techniques and algorithms are crucial in identifying patterns in biological sciences. These strategies have been aiding researchers in categorising medical images and forecasting models to gain a thorough comprehension of intricate medical issues. Deep learning is an area of machine learning that focuses on Artificial Neural Networks (ANNs), which are algorithms designed to mimic the structure and function of the brain. This article utilises a VGG16 model to classify MRI brain images and distinguish between brains affected by PD with normal healthy brains and non-PD with abnormal and unhealthy brains. They are classifying intricate clinical MRI data to identify disorders such as Parkinson's disease or determine the disease's stage with VGG16 Machine Learning Techniques. By employing Machine Learning Techniques, we effectively categorised individuals with Parkinson's disease from those without the condition, achieving an accuracy rate of 95.34% without implementing batch normalisation. Consequently, our research can efficiently employ the same structure to carry out different medical image classification tasks or more intricate systems.