| Popis: |
The prevalence of the neuro generative disease is rapidly increasing in recent years. According to WHO nearly 70 million people suffer due to the brain disorders. The types of brain diseases are Alzheimer Disease, Dementia, Brain Tumor, Epilepsy, Mental Disorders, Parkinson’s disease. Among this Alzheimer disease, Brain Tumor, Parkinson’s Disease and seizure disorders are the most common diseases. The main causes of this diseases are the genetic and environmental factors including diet, smoking and traumatic brain injury, diabetes and other medical diseases contribute to the risk of developing this form of diseases. The main purpose of this work is to develop the computerized brain disease detection method. In this proposed work three brain disease are taken namely Alzheimer, Tumor, Parkinson. The inceptionv3 model and VGG19 are used to detect the brain disease. For efficient detection the transfer learning approach is used. In every deep learning model combined with two set of action one is feature extraction and another one is classification. In this proposed work a novel method is implemented. The deep learning models are used only for the feature extraction purpose. The convolutional features are extracted from the brain images and the Random Forest classifier classify the brain diseases in to Alzheimer, Tumor, Parkinson and Normal brain. Comparison of these the Inceptionv3 with Random Forest outperform well with the accuracy of 95%. |
