Self-paced learning and privileged information based KRR classification algorithm for diagnosis of Parkinson’s disease

• Published in: Neuroscience letters Vol. 766; p. 136312
• Main Authors: Peng, Bo, Gong, Zhenjia, Zhang, Yu, Shen, Bo, Pang, Chunying, Zhang, Li, Dai, Yakang
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier B.V 01.01.2022
• Subjects: Aged; Computer aided diagnosis; Diagnosis, Computer-Assisted - methods; Female; Humans; Image Interpretation, Computer-Assisted - methods; Kernel ridge regression; Machine Learning; Magnetic resonance image; Magnetic Resonance Imaging; Male; Middle Aged; Neural Networks, Computer; Neuroimaging - methods; Parkinson Disease - diagnostic imaging; Parkinson's disease; Privileged information; Self-paced learning; Privileged information; Kernel ridge regression; Parkinson's disease; Computer aided diagnosis; Self-paced learning; Magnetic resonance image
• ISSN: 0304-3940; 1872-7972; 1872-7972
• DOI: 10.1016/j.neulet.2021.136312

•The aim is to improve the CAD accuracy of PD and discover neuroimaging biomarkers.•We proposed a privileged based diagnosis method for PD.•Multiple features are extracted from MRI images and mapping to high-dimensional space.•Different source domain features and target domain features are sent into KRR+ classifier.•The samples are selected adaptively for training through self-paced learning (SPL). Computer aided diagnosis (CAD) methods for Parkinson's disease (PD) can assist clinicians in diagnosis and treatment. Magnetic resonance imaging (MRI) based CAD methods can help reveal structural changes in brain. Classifier is a key component in CAD system, which directly affects the classification performance. Privileged information (PI) can assist to train the classifier by providing additional information, which makes test samples have less error and improves the classification accuracy. In this paper, we proposed a PI based kernel ridge regression plus (KRR+) in diagnosis of PD. Specifically, morphometric features and brain network features are extracted from MRI. Then, empirical kernel mapping feature expression method is used to make the data separable in high-dimensional space. Besides, we introduce self-paced learning that can adaptively select the sample in training of the model, which can further improve the classification performance. The experimental results show that the proposed method is effective for PD diagnosis, its performance is superior to existing classification model. This method is helpful to assist clinicians to find out possible neuroimaging biomarkers in the diagnosis of PD.