101. Classification of the Multiple Stages of Parkinson’s Disease by a Deep Convolution Neural Network Based on 99mTc-TRODAT-1 SPECT Images
- Author
-
Huei-Yung Chen, Shih-Yen Hsu, Tai-Been Chen, Wei-Chang Du, Yi-Chen Wu, Yung-Hui Huang, Li-Ren Yeh, Yun-Hsuan Hsu, Ming-Chia Lin, and Wen-Hung Twan
- Subjects
Male ,Parkinson's disease ,Computer science ,Physics::Medical Physics ,Pharmaceutical Science ,Single-photon emission computed tomography ,Convolutional neural network ,Grayscale ,030218 nuclear medicine & medical imaging ,Analytical Chemistry ,convolution neural network ,0302 clinical medicine ,Drug Discovery ,medicine.diagnostic_test ,Brain ,Technetium ,Parkinson Disease ,Middle Aged ,Quantitative Biology::Genomics ,Chemistry (miscellaneous) ,SPECT ,Molecular Medicine ,Female ,Astrophysics::High Energy Astrophysical Phenomena ,Computer Science::Neural and Evolutionary Computation ,Article ,lcsh:QD241-441 ,03 medical and health sciences ,lcsh:Organic chemistry ,Region of interest ,Spect imaging ,medicine ,Humans ,Physical and Theoretical Chemistry ,Aged ,Retrospective Studies ,Tomography, Emission-Computed, Single-Photon ,Quantitative Biology::Neurons and Cognition ,business.industry ,Deep learning ,Organic Chemistry ,deep learning ,Pattern recognition ,medicine.disease ,Corpus Striatum ,nervous system diseases ,Parkinson’s disease ,Neural Networks, Computer ,Artificial intelligence ,business ,030217 neurology & neurosurgery ,Kappa - Abstract
Single photon emission computed tomography (SPECT) has been employed to detect Parkinson&rsquo, s disease (PD). However, analysis of the SPECT PD images was mostly based on the region of interest (ROI) approach. Due to limited size of the ROI, especially in the multi-stage classification of PD, this study utilizes deep learning methods to establish a multiple stages classification model of PD. In the retrospective study, the 99mTc-TRODAT-1 was used for brain SPECT imaging. A total of 202 cases were collected, and five slices were selected for analysis from each subject. The total number of images was thus 1010. According to the Hoehn and Yahr Scale standards, all the cases were divided into healthy, early, middle, late four stages, and HYS I~V six stages. Deep learning is compared with five convolutional neural networks (CNNs). The input images included grayscale and pseudo color of two types. The training and validation sets were 70% and 30%. The accuracy, recall, precision, F-score, and Kappa values were used to evaluate the models&rsquo, performance. The best accuracy of the models based on grayscale and color images in four and six stages were 0.83 (AlexNet), 0.85 (VGG), 0.78 (DenseNet) and 0.78 (DenseNet).
- Published
- 2020