Utility of Multi-Modal MRI for Differentiating of Parkinson's Disease and Progressive Supranuclear Palsy Using Machine Learning

Abstract

Background: Patients with Parkinson's disease (PD) and progressive supranuclear palsy Richardson's syndrome (PSP-RS) often show overlapping clinical features, leading to misdiagnoses. The objective of this study was to investigate the feasibility and utility of using multi-modal MRI datasets for an automatic differentiation of PD patients, PSP-RS patients, and healthy control (HC) subjects.Material and Methods: T1-weighted, T2-weighted, and diffusion-tensor (DTI) MRI datasets from 45 PD patients, 20 PSP-RS patients, and 38 HC subjects were available for this study. Using an atlas-based approach, regional values of brain morphology (T1-weighted), brain iron metabolism (T2-weighted), and microstructural integrity (DTI) were measured and employed for feature selection and subsequent classification using combinations of various established machine learning methods.Results: The optimal machine learning model using regional morphology features only achieved a classification accuracy of 65% (67/103 correct classifications) differentiating PD patients, PSP-RS patients, and HC subjects. The optimal machine learning model using only quantitative T2 values performed slightly better and achieved an accuracy of 75.7% (78/103). The optimal classifier using DTI features alone performed considerably better with 95.1% accuracy (98/103). The optimal multi-modal classifier using all features also achieved an accuracy of 95.1% but required more features and achieved a slightly lower F1-score compared to the optimal model using DTI features alone.Conclusion: Machine learning models using multi-modal MRI perform significantly better than uni-modal machine learning models using morphological parameters based on T1-weighted MRI datasets alone or brain iron metabolism markers based on T2-weighted MRI datasets alone. However, machine learnig models using regional brain microstructural integrity metrics computed from DTI datasets perform similar to the optimal multi-modal machine learning model. Thus, given the results from this study cohort, it appears that morphology and brain iron metabolism markers may not provide additional value for classification compared to using DTI metrics alone.

Machine Learning for Detecting Parkinson’s Disease by Resting-State Functional Magnetic Resonance Imaging: A Multicenter Radiomics Analysis

“…Both the two datasets had good classification performance, which further indicated the good classification performance and generalization of our model. Third, it has been reported that combining multimodal data and clinical data can improve the performance of the machine learning model (Shi et al, 2021a;Talai et al, 2021), but the primary set in this study only contained ALFF data. A subsequent study should incorporate other modal MRI data, metrics, and clinical data to construct and evaluate the model.…”

“…Previous studies have confirmed the value of rs-fMRI in neuropsychiatric diseases ( Szewczyk-Krolikowski et al, 2014 ; Hu et al, 2015 ; O’Callaghan et al, 2016 ). Recently, with the development of machine learning technologies, more and more studies have used machine learning methods to explore the classification, prognosis prediction, and physiological mechanism of neuropsychiatric diseases, including PD ( Cao et al, 2020 ; Lin et al, 2020 ; Pang et al, 2021 ; Shu et al, 2021 ; Talai et al, 2021 ; Zhang et al, 2021 ). The ROI-based feature extraction is the most commonly used feature extraction method ( Wang L. et al, 2020 ; Zhao et al, 2020 ; Shi et al, 2021b ; Talai et al, 2021 ), and it is a useful method to reduce the data dimensionality ( Wang L. et al, 2020 ).…”

“…Recently, with the development of machine learning technologies, more and more studies have used machine learning methods to explore the classification, prognosis prediction, and physiological mechanism of neuropsychiatric diseases, including PD ( Cao et al, 2020 ; Lin et al, 2020 ; Pang et al, 2021 ; Shu et al, 2021 ; Talai et al, 2021 ; Zhang et al, 2021 ). The ROI-based feature extraction is the most commonly used feature extraction method ( Wang L. et al, 2020 ; Zhao et al, 2020 ; Shi et al, 2021b ; Talai et al, 2021 ), and it is a useful method to reduce the data dimensionality ( Wang L. et al, 2020 ). Functionally defined parcelation and high spatial resolution segmentation might be able to detect a more significant difference, and the anatomical boundary might not match the functional boundary that has been reported in previous literature ( Rosenberg et al, 2016 ; Chen et al, 2018 ).…”

“…Support vector machine is one of the most commonly used machine models, especially in neuroimaging studies in which the sample size is relatively limited ( Hong et al, 2017 ; Tian et al, 2020 ; Shu et al, 2021 ; Talai et al, 2021 ; Zhang et al, 2021 ). SVM incorporates several advantageous properties to reduce overfitting and deliver good generalization performance despite a small sample size ( Hong et al, 2017 ; Mo et al, 2019 ).…”

A Parkinson’s disease patient displaying increased neuromelanin-sensitive areas in the substantia nigra after rehabilitation with tDCS: a case report

“…Second, we assessed only NM-MRI in the present study since we focused on dopaminergic neurons to assess their functional recovery. However, recent studies reported that multimodal MRI assessments including not only NM-MRI, but also structural imaging, nigrosome imaging, iron-sensitive imaging, diffusion tensor imaging (DTI), arterial spin labeling (ASL), and resting-state fMRI to assess functional connectivity would be effective to diagnose PD and understand pathophysiological mechanisms of PD symptoms (Bae et al, 2021;Heim et al, 2017;Talai et al, 2021). Although the present study does not provide information on full pathophysiological changes in the patient, the results suggest that NM-MRI may be useful to assess responses of dopaminergic neurons to treatments.…”