Different patterns of gray matter density in early- and middle-late-onset Parkinson's disease: a voxel-based morphometry study

A more objective PD diagnostic model: integrating texture feature markers of cerebellar gray matter and white matter through machine learning

Objective

The purpose of this study is to explore whether machine learning can be used to establish an effective model for the diagnosis of Parkinson's disease (PD) by using texture features extracted from cerebellar gray matter and white matter, so as to identify subtle changes that cannot be observed by the naked eye.

Method

This study involved a data collection period from June 2010 to March 2023, including 374 subjects from two cohorts. The Parkinson's Progression Markers Initiative (PPMI) served as the training set, with control group and PD patients (HC: 102 and PD: 102) from 24 global sites. Our institution's data was utilized as the test set (HC: 91 and PD: 79). Machine learning was employed to establish multiple models for PD diagnosis based on texture features of the cerebellum's gray and white matter. Results underwent evaluation through 5-fold cross-validation analysis, calculating the area under the receiver operating characteristic curve (AUC) for each model. The performance of each model was compared using the Delong test, and the interpretability of the optimized model was further augmented by employing Shapley additive explanations (SHAP).

Results

The AUCs for all pipelines in the validation dataset were compared using FeAture Explorer (FAE) software. Among the models established by Kruskal-Wallis (KW) and logistic regression via Lasso (LRLasso), the AUC was highest using the "one-standard error" rule. 'WM_original_glrlm_GrayLevelNonUniformity' was considered the most stable and predictive feature.

Conclusion

The texture features of cerebellar gray matter and white matter combined with machine learning may have potential value in the diagnosis of Parkinson's disease, in which the heterogeneity of white matter may be a more valuable imaging marker.

Extracting the most discriminating functional connections in mild traumatic brain injury based on machine learning

BACKGROUND

Mild traumatic brain injury (mTBI) is characterized as brain microstructural damage, which may cause a wide range of brain functional disturbances and emotional problems. Brain network analysis based on machine learning is an important means of neuroimaging research. Obtaining the most discriminating functional connection is of great significance to analyze the pathological mechanism of mTBI.

METHODS

To better obtain the most discriminating features of functional connection networks, this study proposes a hierarchical feature selection pipeline (HFSP) composed of Variance Filtering (VF), Lasso, and Principal Component Analysis (PCA). Ablation experiments indicate that each module plays a positive role in classification, validating the robustness and reliability of the HFSP. Furthermore, the HFSP is compared with recursive feature elimination (RFE), elastic net (EN), and locally linear embedding (LLE), verifying its superiority. In addition, this study also utilizes random forest (RF), SVM, Bayesian, linear discriminant analysis (LDA), and logistic regression (LR) as classifiers to evaluate the generalizability of HFSP.

RESULTS

The results show that the indexes obtained from RF are the highest, with accuracy = 89.74%, precision = 91.26%, recall = 89.74%, and F1 score = 89.42%. The HFSP selects 25 pairs of the most discriminating functional connections, mainly distributed in the frontal lobe, occipital lobe, and cerebellum. Nine brain regions show the largest node degree.

LIMITATIONS

The number of samples is small. This study only includes acute mTBI.

CONCLUSIONS

The HFSP is a useful tool for extracting discriminating functional connections and may contribute to the diagnostic processes.

Large-scale activation likelihood estimation meta-analysis of parkinsonian disorders

Parkinsonism is a feature of several neurodegenerative disorders, including Parkinson's disease, progressive supranuclear palsy, corticobasal syndrome and multiple system atrophy. Neuroimaging studies have yielded insights into parkinsonian disorders; however, due to variability in results, the brain regions consistently implicated in these disorders remain to be characterized. The aim of this meta-analysis was to identify consistent brain abnormalities in individual parkinsonian disorders (Parkinson's disease, progressive supranuclear palsy, corticobasal syndrome and multiple system atrophy) and to investigate any shared abnormalities across disorders. A total of 44 591 studies were systematically screened following searches of two databases. A series of whole-brain activation likelihood estimation meta-analyses were performed on 132 neuroimaging studies (69 Parkinson's disease; 23 progressive supranuclear palsy; 17 corticobasal syndrome; and 23 multiple system atrophy) utilizing anatomical MRI, perfusion or metabolism PET and single-photon emission computed tomography. Meta-analyses were performed in each parkinsonian disorder within each imaging modality, as well as across all included disorders. Results in progressive supranuclear palsy and multiple system atrophy aligned with current imaging markers for diagnosis, encompassing the midbrain, and brainstem and putamen, respectively. PET imaging studies of patients with Parkinson's disease most consistently reported abnormality of the middle temporal gyrus. No significant clusters were identified in corticobasal syndrome. When examining abnormalities shared across all four disorders, the caudate was consistently reported in MRI studies, whilst the thalamus, inferior frontal gyrus and middle temporal gyri were commonly implicated by PET. To our knowledge, this is the largest meta-analysis of neuroimaging studies in parkinsonian disorders and the first to characterize brain regions implicated across parkinsonian disorders.

Analysis of synthetic magnetic resonance images by multi-channel segmentation increases accuracy of volumetry in the putamen and decreases mis-segmentation in the dural sinuses

BACKGROUND

Voxel-based morphometry (VBM) using magnetic resonance imaging (MR) has been used to estimate cortical atrophy associated with various diseases. However, there are mis-segmentations of segmented gray matter image in VBM.

PURPOSE

To study a twofold evaluation of single- and multi-channel segmentation using synthetic MR images: (1) mis-segmentation of segmented gray matter images in transverse and cavernous sinuses; and (2) accuracy and repeatability of segmented gray matter images.

MATERIAL AND METHODS

A total of 13 healthy individuals were scanned with 3D quantification using an interleaved Look-Locker acquisition sequence with a T2 preparation pulse (3D-QALAS) sequence on a 1.5-T scanner. Three of the 13 healthy participants were scanned five consecutive times for evaluation of repeatability. We used SyMRI software to create images with three contrasts: T1-weighted (T1W), T2-weighted (T2W), and proton density-weighted (PDW) images. Manual regions of interest (ROI) on T1W imaging were individually set as the gold standard in the transverse sinus, cavernous sinus, and putamen. Single-channel (T1W) and multi-channel (T1W + T2W, T1W + PDW, and T1W + T2W + PDW imaging) segmentations were performed with statistical parametric mapping 12 software.

RESULTS

We found that mis-segmentations in both the transverse and cavernous sinuses were large in single-channel segmentation compared with multi-channel segmentations. Furthermore, the accuracy of segmented gray matter images in the putamen was high in both multi-channel T1W + PDW and T1W + T2W + PDW segmentations compared with other segmentations. Finally, the highest repeatability of left putamen volumetry was found with multi-channel segmentation T1WI + PDWI.

CONCLUSION

Multi-channel segmentation with T1WI + PDWI provides good results for VBM compared with single-channel and other multi-channel segmentations.

Abnormal intra- and inter-network functional connectivity of brain networks in early-onset Parkinson's disease and late-onset Parkinson's disease

Objective

The purpose of this study is to look into the altered functional connectivity of brain networks in Early-Onset Parkinson's Disease (EOPD) and Late-Onset Parkinson's Disease (LOPD), as well as their relationship to clinical symptoms.

Methods

A total of 50 patients with Parkinson' disease (28 EOPD and 22 LOPD) and 49 healthy controls (25 Young Controls and 24 Old Controls) were admitted to our study. Employing independent component analysis, we constructed the brain networks of EOPD and Young Controls, LOPD and Old Controls, respectively, and obtained the functional connectivity alterations in brain networks.

Results

Cerebellar network (CN), Sensorimotor Network (SMN), Executive Control Network (ECN), and Default Mode Network (DMN) were selected as networks of interest. Compared with their corresponding health controls, EOPD showed increased functional connectivity within the SMN and ECN and no abnormalities of inter-network functional connectivity were found, LOPD demonstrated increased functional connectivity within the ECN while decreased functional connectivity within the CN. Furthermore, in LOPD, functional connectivity between the SMN and DMN was increased. The functional connectivity of the post-central gyrus within the SMN in EOPD was inversely correlated with the Unified Parkinson's Disease Rating Scale Part III scores. Age, age of onset, and MMSE scores are significantly different between EOPD and LOPD (p < 0.05).

Conclusion

There is abnormal functional connectivity of networks in EOPD and LOPD, which could be the manifestation of the associated pathological damage or compensation.

Structural and functional abnormalities in Parkinson's disease based on voxel-based morphometry and resting-state functional magnetic resonance imaging

OBJECTIVE

To explore differences in gray matter volume (GMV) and white matter volume (WMV) between patients with Parkinson's disease (PD) and healthy controls, and to examine whether the structural abnormalities correlate with functional abnormalities.

METHODS

T1-weighted magnetic resonance imaging and resting-state functional magnetic resonance imaging (fMRI) were performed on 180 patients with PD and 58 age- and sex-matched healthy controls. We used voxel-based morphometry (VBM) to compare GMV and WMV between groups, and resting-state fMRI to compare amplitudes of low-frequency fluctuations (ALFF) in the structurally abnormal brain regions.

RESULTS

Structural neuroimaging showed smaller whole-brain GMV, but not WMV, in patients. Furthermore, VBM revealed smaller GMV in the right superior temporal gyrus (STG) and left frontotemporal space in patients, after correction for multiple comparisons. Patients also showed significantly higher ALFF in the right STG. GMV in the right STG and left frontotemporal space in patients correlated negatively with age and scores on Part III of the Movement Disorder Society Unified Parkinson's Disease Rating Scale, but not with PD duration.

CONCLUSIONS

Structural atrophy in the frontotemporal lobe may be a useful imaging biomarker in PD, such as for detecting disease progression. Furthermore, this structural atrophy appears to correlate with enhanced spontaneous brain activity. This study associates particular structural and functional abnormalities with PD neuropathology.

Brain Activity in Different Brain Areas of Patients With Dry Eye During the Female Climacteric Period According to Voxel-Based Morphometry

We aim to investigate potential morphological alterations of the brain in female climacteric patients with dry eye (DE) and their relationship to behavioral performances. Twenty-five female patients with DE disease during the female climacteric period and 25 age and education-matched healthy controls (HCs) underwent magnetic resonance imaging. Imaging data were analyzed using voxel-based morphometry (VBM) to identify group differences in DE patients and HCs. Compared with HCs, patients with DE during the female climacteric period had significantly decreased VBM in the Putamen_L, Thalamus_R, Precuneus_L, Frontal_Sup_R, Cingulum_Mid_L, and Frontal_Mid_L. There was increased VBM in the Temporal_Pole_Sup_R, Precentral_R and Insula_L. Receiver operating characteristic curve analysis indicated that the VBM method has clear potential for diagnosis of DE patients during the climacteric period. Correlation analysis found a negative correlation between the VBM values of the Putamen_L and the anxiety score (AS) and depression score (DS), a positive correlation was found between VBM values of the Temporal_Pole_Sup_R and AS. Moreover, VBM values in the Cingulum_Mid_L were positively correlated with AS and DS. These results revealed abnormal spontaneous activity in the brain regions of patients with DE during the climacteric period, which may indicate underlying pathological mechanisms. These results may help to advance clinical treatments.

Application of Nanomaterials in Neurodegenerative Diseases

Neurodegenerative diseases, including Alzheimer's disease, Parkinson's disease, and Huntington's disease are very harmful brain lesions. Due to the difficulty in obtaining therapeutic drugs, the best treatment for neurodegenerative diseases is often not available. In addition, the bloodbrain barrier can effectively prevent the transfer of cells, particles and macromolecules (such as drugs) in the brain, resulting in the failure of the traditional drug delivery system to provide adequate cellular structure repair and connection modes, which are crucial for the functional recovery of neurodegenerative diseases. Nanomaterials are designed to carry drugs across the blood-brain barrier for targets. Nanotechnology uses engineering materials or equipment to interact with biological systems at the molecular level to induce physiological responses through stimulation, response and target site interactions, while minimizing the side effects, thus revolutionizing the treatment and diagnosis of neurodegenerative diseases. Some magnetic nanomaterials play a role as imaging agents or nanoprobes for Magnetic Resonance Imaging to assist in the diagnosis of neurodegenerative diseases. Although the current research on nanomaterials is not as useful as expected in clinical applications, it achieves a major breakthrough and guides the future development direction of nanotechnology in the application of neurodegenerative diseases. This review briefly discusses the application and advantages of nanomaterials in neurodegenerative diseases. Data for this review were identified by searches of PubMed, and references from relevant articles published in English between 2015 and 2019 using the search terms "nanomaterials", "neurodegenerative diseases" and "blood-brain barrier".

Increased Plasma Heme Oxygenase-1 Levels in Patients With Early-Stage Parkinson's Disease

Introduction: Heme oxygenase-1 (HO-1) is a 32 kDa stress-response protein implicated in the pathogenesis of Parkinson’s disease (PD). Biliverdin is derived from heme through a reaction mediated by HO-1 and protects cells from oxidative stress. However, iron and carbon monoxide produced by the catabolism of HO-1 exert detrimental effects on patients with PD. The purpose of this study was to determine whether plasma HO-1 levels represent a biomarker of PD and to further explore the underlying mechanism of increased HO-1 levels by applying voxel-based morphometry (VBM).Methods: We measured plasma HO-1 levels using an enzyme-linked immunosorbent assay (ELISA) in 156 subjects, including 81 patients with early- and advanced-stage PD and 75 subjects without PD. The analyses were adjusted to control for confounders such as age, sex, and medication. We analyzed T1-weighted magnetic resonance imaging (MRI) data from 74 patients with PD using VBM to elucidate the association between altered brain volumes and HO-1 levels. Then, we compared performance on MMSE sub-items between PD patients with low and high levels of HO-1 using Mann-Whitney U tests.Results: Plasma HO-1 levels were significantly elevated in PD patients, predominantly those with early-stage PD, compared with controls (p < 0.05). The optimal cutoff value for patients with early PD was 2.245 ng/ml HO-1 [area under the curve (AUC) = 0.654]. Plasma HO-1 levels were unaffected by sex, age, and medications (p > 0.05). The right hippocampal volume was decreased in the subset of PD patients with high HO-1 levels (p < 0.05). A weak correlation was observed between right hippocampal volume and plasma HO-1 levels (r = −0.273, p = 0.018). There was no difference in total MMSE scores between the low- and high-HO-1 groups (p > 0.05), but the high-HO-1 group had higher language scores than the low-HO-1 group (p < 0.05).Conclusions: Plasma HO-1 levels may be a promising biomarker of early PD. Moreover, a high plasma concentration of the HO-1 protein is associated with a reduction in right hippocampal volume.

Social Cognition in Patients with Early-Onset Parkinson's Disease

Social cognition (SC) deficits have been linked to Parkinson's disease (PD) but have been less well researched than general cognitive processes, especially in early-onset PD (EOPD), despite this population often having greater social and family demands. Most studies focus on recognition of facial emotion, theory of mind (ToM), and decision-making domains, with limited research reporting on social reasoning. The main objective of this work was to compare SC ability across four domains: emotional processing, social reasoning, ToM, and decision-making between patients with EOPD and healthy controls. Twenty-five nondemented patients with EOPD and 25 controls matched for sex, age, and educational level were enrolled. A battery that included six SC tests was administered to all study participants; a decision-making scale was completed by participants' partners. Statistically significant differences were found between patients with EOPD and controls in all subtests across the four SC domains studied. The EOPD group demonstrated worse performance on all tasks, with large effect sizes. Differences remained significant after adjusting for Montreal Cognitive Assessment (MoCA) test scores for all SC subtests except the decision-making scale and the Iowa gambling task. No significant correlations between SC and other clinical PD variables were found. Our study shows that patients with EOPD perform significantly below controls in multiple SC domains affecting recognition of facial emotion, social reasoning, ToM, and decision-making. Only decision-making seems to be mediated by overall cognitive ability. The confounding or contributing effect of other clinical PD variables should be studied further.

Distinct Changes in Global Brain Synchronization in Early-Onset vs. Late-Onset Parkinson Disease

Early- and late-onset Parkinson's disease (EOPD and LOPD, respectively) have different risk factors, clinical features, and disease course; however, the functional outcome of these differences have not been well characterized. This study investigated differences in global brain synchronization changes and their clinical significance in EOPD and LOPD patients. Patients with idiopathic PD including 25 EOPD and 24 LOPD patients, and age- and sex-matched healthy control (HC) subjects including 27 younger and 26 older controls (YCs and OCs, respectively) were enrolled. Voxel-based degree centrality (DC) was calculated as a measure of global synchronization and compared between PD patients and HC groups matched in terms of disease onset and severity. DC was decreased in bilateral Rolandic operculum and left insula and increased in the left superior frontal gyrus (SFG) and precuneus of EOPD patients compared to YCs. DC was decreased in the right putamen, mid-cingulate cortex, bilateral Rolandic operculum, and left insula and increased in the right cerebellum-crus1 of LOPD patients compared to OCs. Correlation analyses showed that DC in the right cerebellum-crus1 was inversely associated with the Hamilton Depression Scale (HDS) score in LOPD patients. Thus, EOPD and LOPD patients show distinct alterations in global synchronization relative to HCs. Furthermore, our results suggest that the left SFG and right cerebellum-crus1 play important roles in the compensation for corticostriatal-thalamocortical loop injury in EOPD and LOPD patients, whereas the cerebellum is a key hub in the neural mechanisms underlying LOPD with depression. These findings provide new insight into the clinical heterogeneity of the two PD subtypes.

Altered Functional Network Associated With Cognitive Performance in Early Parkinson Disease Measured by Eigenvector Centrality Mapping

Objective: To investigate relationships between whole-brain functional changes and the performance of multiple cognitive functions in early Parkinson’s disease (PD).

Methods: In the current study, we evaluated resting-state functional MRI (rsfMRI) data and neuropsychological assessments for various cognitive functions in a cohort with 84 early PD patients from the Parkinson’s Progression Markers Initiative (PPMI). Eigenvector centrality (EC) mapping based on rsfMRI was used to identify the functional connectivity of brain areas correlated with different neuropsychological scores at a whole-brain level.

Results: Our study demonstrated that in the early PD patients, scores of Letter Number Sequencing (LNS) were positively correlated with EC in the left inferior occipital gyrus (IOG) and lingual gyrus. The immediate recall scores of Hopkins Verbal Learning Test-Revised (HVLT-R) were positively correlated with EC in the left superior frontal gyrus. No correlation was found between the EC and other cognitive performance scores.

Conclusions: Functional alternations in the left occipital lobe (inferior occipital and lingual gyrus) and left superior frontal gyrus may account for the performance of working memory and immediate recall memory, respectively in early PD. These results may broaden the understanding of the potential mechanism of cognitive impairments in early PD.

Use of Magnetic Resonance Imaging and Artificial Intelligence in Studies of Diagnosis of Parkinson's Disease

Parkinson's disease (PD) is a common neurodegenerative disorder. It has a delitescent onset and a slow progress. The clinical manifestations of PD in patients are highly heterogeneous. Thus, PD diagnosis process is complex and mainly depends on the professional knowledge and experience of the physician. Magnetic resonance imaging (MRI) could detect the small changes in the brain of PD patients, and quantitative analysis of brain MRI may improve the clinical diagnosis efficiency. However, due to the complexity of clinical courses in PD and the high dimensionality in multimodal MRI data, traditional mathematical analysis could not effectively extract the huge information in them. Up to now, the accuracy of PD diagnosis in large sample size is still unsatisfying. As artificial intelligence (AI) is becoming more mature, varieties of statistical models and machine learning (ML) algorithms have been used for quantitative imaging data analysis to explore a diagnostic result. This review aims to state an overview of existing research recently that used statistical ML/AI methods to perform quantitative analysis of MR image data for the study of PD diagnosis. First we review the recent research in three subareas: diagnosis, differential diagnosis, and subtyping of PD. Then we described the overall workflow from MR image to classification result. Finally, we summarized a critical assessment of the current research and provide some recommendations for likely future research developments and trends.