Treffer: Zebra based optimal deep learning for Parkinson's disease detection using brain MRI images.

Parkinson's disease (PD) is reflected to be the second most serious neurological disease in the world after Alzheimer's disease. Early diagnosis of PD is crucial to decrease complications. Existing studies have developed various methods to effectively detect PD. However, no appropriate medical tests or methods are available for conclusive diagnosis in the early stages due to limitations such as higher computational complexity, noisy samples, excessive irrelevant features, etc. The Optimization-assisted Deep Learning (OAssis-DL) model is proposed to mitigate such problems in PD detection. The initial phase is pre-processing the raw input images through the Adaptive Frost Filtering (AFF) model. After pre-processing, feature extraction is performed using Extended Local Optimal Oriented Descriptor (ELOOP) and Discrete Wavelet Transform (DWletT). Next, the optimal features are chosen by the Zebra Optimization Algorithm (ZOA) to reduce the dimensionality problem. The Attention-based Deep-Stack Convolutional Hunter-Prey Network (ADeepSCHN) is then selected as a classifier to detect the PD classes. The OAssis-DL model is trained and tested using the publicly accessible Parkinson's Progression Marker Initiative (PPMI) dataset. The Python platform is used for implementation, and various performance indicators are analyzed based on existing studies. As a result, the OAssis-DL model achieves an overall accuracy of 99.61% respectively. [ABSTRACT FROM AUTHOR]

Copyright of Multimedia Tools & Applications is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)