Stage-dependent differential impact of network communication on cognitive function across the continuum of cognitive decline in Parkinson's disease

OBJECTIVE

Our objective was to explore the patterns of resting-state network (RSN) connectivity alterations and investigate how the influences of individual-level network connections on cognition varied across clinical stages without assuming a constant relationship.

METHODS

108 PD patients with continuum of cognitive decline (PD-NC = 46, PD-MCI = 43, PDD = 19) and 34 healthy controls (HCs) underwent resting-state functional MRI and neuropsychological tests. Independent component analysis (ICA) and graph theory analyses (GTA) were employed to explore RSN connection changes. Additionally, stage-dependent differential impact of network communication on cognitive performance were examined using sparse varying coefficient modeling.

RESULTS

Compared to HCs, the dorsal attention network (DAN) and dorsal sensorimotor network (dSMN) were central networks with decreased connections in PD-NC and PD-MCI stage, while the lateral visual network (LVN) emerged as a central network in patients with dementia. Additionally, connectivity of the cerebellum network (CBN) increased in the PD-NC and PD-MCI stages. GTA demonstrated decreased nodal metrics for DAN and dSMN, coupled with an increase for CBN. Moreover, the degree centrality (DC) values of DAN and dSMN exhibited a stage-dependent differential impact on cognitive performance across the continuum of cognitive decline.

CONCLUSION

Our findings suggest that across the progression of cognitive impairment, the LVN gradually transitions into a core node with reduced connectivity, while the enhancement of connections in CBN diminishes. Furthermore, the non-linear relationship between the DC values of RSNs and cognitive decline indicates the potential for tailored interventions targeting specific stages.

Does the Montreal Cognitive Assessment (MoCA) identify cognitive impairment profiles in Parkinson's disease? An exploratory study

An important proportion of patients with Parkinson's Disease (PD) present signs of cognitive impairment, although this is heterogeneous. In an attempt to classify this, the dual syndrome hypothesis distinguishes between two profiles: one defined by attentional and executive problems with damage in anterior cerebral regions, and another with mnesic and visuospatial alterations, with damage in posterior cerebral regions. The Montreal Cognitive Assessment (MoCA) is one of the recommended screening tools, and one of the most used, to assess cognitive impairment in PD. However, its ability to specifically identify these two profiles of cognitive impairment has not been studied. The aim of this study was, therefore, to analyze the capacity of the MoCA to detect cognitive impairment, and also to identify anterior and posterior profiles defined by the dual syndrome hypothesis. For this purpose, 59 patients with idiopathic PD were studied with the MoCA and a neuropsychological battery of tests covering all cognitive domains. Results of logistic regression analysis with ROC (Receiver Operating Characteristic) curves showed that MoCA detected cognitive impairment and identified patients with a profile of anterior/attentional and executive deficit, with acceptable sensibility and specificity. However, it did not identify patients with a posterior/mnesic-visuospatial impairment. We discuss the reasons for the lack of sensitivity of MoCA in this profile, and other possible implications of these results with regards the usefulness of this tool to assess cognitive impairment in PD.

Identifying Biases in a Multicenter MRI Database for Parkinson's Disease Classification: Is the Disease Classifier a Secret Site Classifier?

Sharing multicenter imaging datasets can be advantageous to increase data diversity and size but may lead to spurious correlations between site-related biological and non-biological image features and target labels, which machine learning (ML) models may exploit as shortcuts. To date, studies analyzing how and if deep learning models may use such effects as a shortcut are scarce. Thus, the aim of this work was to investigate if site-related effects are encoded in the feature space of an established deep learning model designed for Parkinson's disease (PD) classification based on T1-weighted MRI datasets. Therefore, all layers of the PD classifier were frozen, except for the last layer of the network, which was replaced by a linear layer that was exclusively re-trained to predict three potential bias types (biological sex, scanner type, and originating site). Our findings based on a large database consisting of 1880 MRI scans collected across 41 centers show that the feature space of the established PD model (74% accuracy) can be used to classify sex (75% accuracy), scanner type (79% accuracy), and site location (71% accuracy) with high accuracies despite this information never being explicitly provided to the PD model during original training. Overall, the results of this study suggest that trained image-based classifiers may use unwanted shortcuts that are not meaningful for the actual clinical task at hand. This finding may explain why many image-based deep learning models do not perform well when applied to data from centers not contributing to the training set.

Exploiting macro- and micro-structural brain changes for improved Parkinson's disease classification from MRI data

Parkinson's disease (PD) is the second most common neurodegenerative disease. Accurate PD diagnosis is crucial for effective treatment and prognosis but can be challenging, especially at early disease stages. This study aimed to develop and evaluate an explainable deep learning model for PD classification from multimodal neuroimaging data. The model was trained using one of the largest collections of T1-weighted and diffusion-tensor magnetic resonance imaging (MRI) datasets. A total of 1264 datasets from eight different studies were collected, including 611 PD patients and 653 healthy controls (HC). These datasets were pre-processed and non-linearly registered to the MNI PD25 atlas. Six imaging maps describing the macro- and micro-structural integrity of brain tissues complemented with age and sex parameters were used to train a convolutional neural network (CNN) to classify PD/HC subjects. Explainability of the model's decision-making was achieved using SmoothGrad saliency maps, highlighting important brain regions. The CNN was trained using a 75%/10%/15% train/validation/test split stratified by diagnosis, sex, age, and study, achieving a ROC-AUC of 0.89, accuracy of 80.8%, specificity of 82.4%, and sensitivity of 79.1% on the test set. Saliency maps revealed that diffusion tensor imaging data, especially fractional anisotropy, was more important for the classification than T1-weighted data, highlighting subcortical regions such as the brainstem, thalamus, amygdala, hippocampus, and cortical areas. The proposed model, trained on a large multimodal MRI database, can classify PD patients and HC subjects with high accuracy and clinically reasonable explanations, suggesting that micro-structural brain changes play an essential role in the disease course.

A multi-center distributed learning approach for Parkinson's disease classification using the traveling model paradigm

Distributed learning is a promising alternative to central learning for machine learning (ML) model training, overcoming data-sharing problems in healthcare. Previous studies exploring federated learning (FL) or the traveling model (TM) setup for medical image-based disease classification often relied on large databases with a limited number of centers or simulated artificial centers, raising doubts about real-world applicability. This study develops and evaluates a convolution neural network (CNN) for Parkinson's disease classification using data acquired by 83 diverse real centers around the world, mostly contributing small training samples. Our approach specifically makes use of the TM setup, which has proven effective in scenarios with limited data availability but has never been used for image-based disease classification. Our findings reveal that TM is effective for training CNN models, even in complex real-world scenarios with variable data distributions. After sufficient training cycles, the TM-trained CNN matches or slightly surpasses the performance of the centrally trained counterpart (AUROC of 83% vs. 80%). Our study highlights, for the first time, the effectiveness of TM in 3D medical image classification, especially in scenarios with limited training samples and heterogeneous distributed data. These insights are relevant for situations where ML models are supposed to be trained using data from small or remote medical centers, and rare diseases with sparse cases. The simplicity of this approach enables a broad application to many deep learning tasks, enhancing its clinical utility across various contexts and medical facilities.

Interactions between functional networks in Parkinson's disease mild cognitive impairment

The study of mild cognitive impairment (MCI) is critical to understand the underlying processes of cognitive decline in Parkinson's disease (PD). Functional connectivity (FC) disruptions in PD-MCI patients have been observed in several networks. However, the functional and cognitive changes associated with the disruptions observed in these networks are still unclear. Using a data-driven methodology based on independent component analysis, we examined differences in FC RSNs among PD-MCI, PD cognitively normal patients (PD-CN) and healthy controls (HC) and studied their associations with cognitive and motor variables. A significant difference was found between PD-MCI vs PD-CN and HC in a FC-trait comprising sensorimotor (SMN), dorsal attention (DAN), ventral attention (VAN) and frontoparietal (FPN) networks. This FC-trait was associated with working memory, memory and the UPDRS motor scale. SMN involvement in verbal memory recall may be related with the FC-trait correlation with memory deficits. Meanwhile, working memory impairment may be reflected in the DAN, VAN and FPN interconnectivity disruptions with the SMN. Furthermore, interactions between the SMN and the DAN, VAN and FPN network reflect the intertwined decline of motor and cognitive abilities in PD-MCI. Our findings suggest that the memory impairments observed in PD-MCI are associated with reduced FC within the SMN and between SMN and attention networks.

Image-encoded biological and non-biological variables may be used as shortcuts in deep learning models trained on multisite neuroimaging data

OBJECTIVE

This work investigates if deep learning (DL) models can classify originating site locations directly from magnetic resonance imaging (MRI) scans with and without correction for intensity differences.

MATERIAL AND METHODS

A large database of 1880 T1-weighted MRI scans collected across 41 sites originally for Parkinson's disease (PD) classification was used to classify sites in this study. Forty-six percent of the datasets are from PD patients, while 54% are from healthy participants. After preprocessing the T1-weighted scans, 2 additional data types were generated: intensity-harmonized T1-weighted scans and log-Jacobian deformation maps resulting from nonlinear atlas registration. Corresponding DL models were trained to classify sites for each data type. Additionally, logistic regression models were used to investigate the contribution of biological (age, sex, disease status) and non-biological (scanner type) variables to the models' decision.

RESULTS

A comparison of the 3 different types of data revealed that DL models trained using T1-weighted and intensity-harmonized T1-weighted scans can classify sites with an accuracy of 85%, while the model using log-Jacobian deformation maps achieved a site classification accuracy of 54%. Disease status and scanner type were found to be significant confounders.

DISCUSSION

Our results demonstrate that MRI scans encode relevant site-specific information that models could use as shortcuts that cannot be removed using simple intensity harmonization methods.

CONCLUSION

The ability of DL models to exploit site-specific biases as shortcuts raises concerns about their reliability, generalization, and deployability in clinical settings.

Explainable classification of Parkinson's disease using deep learning trained on a large multi-center database of T1-weighted MRI datasets

INTRODUCTION

Parkinson's disease (PD) is a severe neurodegenerative disease that affects millions of people. Early diagnosis is important to facilitate prompt interventions to slow down disease progression. However, accurate PD diagnosis can be challenging, especially in the early disease stages. The aim of this work was to develop and evaluate a robust explainable deep learning model for PD classification trained from one of the largest collections of T1-weighted magnetic resonance imaging datasets.

MATERIALS AND METHODS

A total of 2,041 T1-weighted MRI datasets from 13 different studies were collected, including 1,024 datasets from PD patients and 1,017 datasets from age- and sex-matched healthy controls (HC). The datasets were skull stripped, resampled to isotropic resolution, bias field corrected, and non-linearly registered to the MNI PD25 atlas. The Jacobian maps derived from the deformation fields together with basic clinical parameters were used to train a state-of-the-art convolutional neural network (CNN) to classify PD and HC subjects. Saliency maps were generated to display the brain regions contributing the most to the classification task as a means of explainable artificial intelligence.

RESULTS

The CNN model was trained using an 85%/5%/10% train/validation/test split stratified by diagnosis, sex, and study. The model achieved an accuracy of 79.3%, precision of 80.2%, specificity of 81.3%, sensitivity of 77.7%, and AUC-ROC of 0.87 on the test set while performing similarly on an independent test set. Saliency maps computed for the test set data highlighted frontotemporal regions, the orbital-frontal cortex, and multiple deep gray matter structures as most important.

CONCLUSION

The developed CNN model, trained on a large heterogenous database, was able to differentiate PD patients from HC subjects with high accuracy with clinically feasible classification explanations. Future research should aim to investigate the combination of multiple imaging modalities with deep learning and on validating these results in a prospective trial as a clinical decision support system.

The assessment of executive functions to test the integrity of the nigrostriatal network: A pilot study

Background

Parkinson’s disease (PD) is a chronic neurodegenerative disorder characterized by motor and non-motor symptoms. The latter mainly include affective, sleep, and cognitive deficits. Non-demented PD patients often demonstrate impairments in several executive domains following neuropsychological evaluation. The current pilot study aims at assessing the discriminatory power of the Frontal Assessment Battery-15 (FAB15) in differentiating (i) non-demented PD patients and healthy controls and (ii) PD patients with more and less pronounced motor symptoms.

Methods

Thirty-nine non-demented early-stage PD patients in the “on” dopamine state (26 females, mean age = 64.51 years, SD = 6.47, mean disease duration = 5.49 years, SD = 2.28) and 39 healthy participants (24 females, mean age = 62.60 years, SD = 5.51) were included in the study. All participants completed the FAB15. Motor symptoms of PD patients were quantified via the Unified Parkinson’s Disease Rating Scale-Part III (UPDRS-Part III) and Hoehn and Yahr staging scale (H&Y).

Results

The FAB15 score, adjusted according to normative data for sex, age, and education, proved to be sufficiently able to discriminate PD patients from healthy controls (AUC = 0.69 [95% CI 0.60–0.75], SE = 0.06, p = 0.04, optimal cutoff = 11.29). Conversely, the battery lacked sufficient discriminative capability to differentiate PD patients based on the severity of motor symptoms.

Conclusion

The FAB15 may be a valid tool for distinguishing PD patients from healthy controls. However, it might be less sensitive in identifying clinical phenotypes characterized by visuospatial impairments resulting from posteroparietal and/or temporal dysfunctions. In line with previous evidence, the battery demonstrated to be not expendable in the clinical practice for monitoring the severity of PD-related motor symptoms.

Identification of metabolic correlates of mild cognitive impairment in Parkinson's disease using magnetic resonance spectroscopic imaging and machine learning

OBJECTIVE

To investigate metabolic changes of mild cognitive impairment in Parkinson's disease (PD-MCI) using proton magnetic resonance spectroscopic imaging (¹H-MRSI).

METHODS

Sixteen healthy controls (HC), 26 cognitively normal Parkinson's disease (PD-CN) patients, and 34 PD-MCI patients were scanned in this prospective study. Neuropsychological tests were performed, and three-dimensional ¹H-MRSI was obtained at 3 T. Metabolic parameters and neuropsychological test scores were compared between PD-MCI, PD-CN, and HC. The correlations between neuropsychological test scores and metabolic intensities were also assessed. Supervised machine learning algorithms were applied to classify HC, PD-CN, and PD-MCI groups based on metabolite levels.

RESULTS

PD-MCI had a lower corrected total N-acetylaspartate over total creatine ratio (tNAA/tCr) in the right precentral gyrus, corresponding to the sensorimotor network (p = 0.01), and a lower tNAA over myoinositol ratio (tNAA/mI) at a part of the default mode network, corresponding to the retrosplenial cortex (p = 0.04) than PD-CN. The HC and PD-MCI patients were classified with an accuracy of 86.4% (sensitivity = 72.7% and specificity = 81.8%) using bagged trees.

CONCLUSION

¹H-MRSI revealed metabolic changes in the default mode, ventral attention/salience, and sensorimotor networks of PD-MCI patients, which could be summarized mainly as 'posterior cortical metabolic changes' related with cognitive dysfunction.

Systematic review of data-driven cognitive subtypes in Parkinson disease

BACKGROUND AND PURPOSE

Recent application of the mild cognitive impairment concept to Parkinson disease (PD) has proven valuable in identifying patients at risk of dementia. However, it has sparked controversy regarding the existence of cognitive subtypes. The present review evaluates the current literature pertaining to data-driven subtypes of cognition in PD.

METHODS

Following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, systematic literature searches for peer-reviewed articles on the topic of cognitive subtyping in PD were performed.

RESULTS

Twenty-two relevant articles were identified in the systematic search. Subtype structures showed either a spectrum of severity or specific domains of impairment. Domain-specific subtypes included amnestic/nonamnestic, memory/executive, and frontal/posterior dichotomies, as well as more complex structures with less definitive groupings. Preliminary longitudinal evidence showed some differences in cognitive progression among subtypes. Neuroimaging evidence provided insight into distinct patterns of brain alterations among subtypes.

CONCLUSIONS

Recurring phenotypes in the literature suggest strong clinical relevance of certain cognitive subtypes in PD. Although the current literature is limited, it raises critical questions about the utility of data-driven methods in cognitive research. The results encourage further integration of neuroimaging research to define the latent neural mechanisms behind divergent subtypes. Although there is no consensus, there appears to be growing consistency and inherent value in identifying cognitive subtypes in PD.

Abnormal static and dynamic functional network connectivity in stable chronic obstructive pulmonary disease

Objective

Many studies have explored the neural mechanisms of cognitive impairment in chronic obstructive pulmonary disease (COPD) patients using the functional MRI. However, the dynamic properties of brain functional networks are still unclear. The purpose of this study was to explore the changes in dynamic functional network attributes and their relationship with cognitive impairment in stable COPD patients.

Materials and methods

The resting-state functional MRI and cognitive assessments were performed on 19 stable COPD patients and 19 age-, sex-, and education-matched healthy controls (HC). We conducted the independent component analysis (ICA) method on the resting-state fMRI data, and obtained seven resting-state networks (RSNs). After that, the static and dynamic functional network connectivity (sFNC and dFNC) were respectively constructed, and the differences of functional connectivity (FC) were compared between the COPD patients and the HC groups. In addition, the correlation between the dynamic functional network attributes and cognitive assessments was analyzed in COPD patients.

Results

Compared to HC, there were significant differences in sFNC among COPD patients between and within networks. COPD patients showed significantly longer mean dwell time and higher fractional windows in weaker connected State I than that in HC. Besides, in comparison to HC, COPD patients had more extensive abnormal FC in weaker connected State I and State IV, and less abnormal FC in stronger connected State II and State III, which were mainly located in the default mode network, executive control network, and visual network. In addition, the dFNC properties including mean dwell time and fractional windows, were significantly correlated with some essential clinical indicators such as FEV₁, FEV₁/FVC, and c-reactive protein (CRP) in COPD patients.

Conclusion

These findings emphasized the differences in sFNC and dFNC of COPD patients, which provided a new perspective for understanding the cognitive neural mechanisms, and these indexes may serve as neuroimaging biomarkers of cognitive performance in COPD patients.

Alterations in degree centrality and cognitive function in breast cancer patients after chemotherapy

The goal of this study was to determine the presence or absence of persistent functional impairments in specific brain regions in breast cancer patients during the recovery period after chemotherapy. We calculated degree centrality (DC) and explored the correlation between brain changes and cognitive scores in 29 female patients with breast cancer who had completed chemotherapy within 1-6 years (C + group) and in 28 age-matched patients with breast cancer who did not receive chemotherapy (C- group). All patients underwent rs-fMRI and cognitive testing. Differences in brain functional activity were explored using DC parameters. Correlations between brain features and cognitive scores were analyzed via correlation analysis. Compared with the C- group, the C + group obtained significantly lower motor and cognitive subscores on the Fatigue Scale for Motor and Cognitive Functions and four subscale scores of the Functional Assessment of Cancer Therapy-Cognitive Function (P < 0.05). Furthermore, the C + group exhibited a significantly higher DC z-score (zDC) in the right superior temporal gyrus and left postcentral gyrus (P < 0.01, FWE-corrected), and a lower zDC in the left caudate nucleus (P < 0.01, FWE-corrected). We found a positive correlation between digit symbol test (DST) scores and zDC values in the right superior temporal gyrus (r = 0.709, P < 0.001), and a negative correlation between DST scores and zDC values in the right angular gyrus (r = -0.784, P < 0.001) and left superior parietal gyrus (r = -0.739, P < 0.001). Chemotherapy can cause abnormal brain activity and cognitive decline in patients with breast cancer, and these effects are likely to persist. DC can be used as an imaging marker for chemotherapy-related cognitive impairment after chemotherapy in breast cancer patients.

Deep Brain Stimulation Modulates Multiple Abnormal Resting-State Network Connectivity in Patients With Parkinson’s Disease

Background

Deep brain stimulation (DBS) improves motor and non-motor symptoms in patients with Parkinson’s disease (PD). Researchers mainly investigated the motor networks to reveal DBS mechanisms, with few studies extending to other networks. This study aimed to investigate multi-network modulation patterns using DBS in patients with PD.

Methods

Twenty-four patients with PD underwent 1.5 T functional MRI (fMRI) scans in both DBS-on and DBS-off states, with twenty-seven age-matched healthy controls (HCs). Default mode, sensorimotor, salience, and left and right frontoparietal networks were identified by using the independent component analysis. Power spectra and functional connectivity of these networks were calculated. In addition, multiregional connectivity was established from 15 selected regions extracted from the abovementioned networks. Comparisons were made among groups. Finally, correlation analyses were performed between the connectivity changes and symptom improvements.

Results

Compared with HCs, PD-off showed abnormal power spectra and functional connectivity both within and among these networks. Some of the abovementioned abnormalities could be corrected by DBS, including increasing the power spectra in the sensorimotor network and modulating the parts of the ipsilateral functional connectivity in different regions centered in the frontoparietal network. Moreover, the DBS-induced functional connectivity changes were correlated with motor and depression improvements in patients with PD.

Conclusion

DBS modulated the abnormalities in multi-networks. The functional connectivity alterations were associated with motor and psychiatric improvements in PD. This study lays the foundation for large-scale brain network research on multi-network DBS modulation.

Resting-State Functional Connectivity in Frontostriatal and Posterior Cortical Subtypes in Parkinson's Disease-Mild Cognitive Impairment

BACKGROUND

The "dual syndrome hypothesis" distinguished two subtypes in mild cognitive impairment (MCI) in Parkinson's disease: frontostriatal, characterized by attentional and executive deficits; and posterior cortical, characterized by visuospatial, memory, and language deficits.

OBJECTIVE

The aim was to identify resting-state functional modifications associated with these subtypes.

METHODS

Ninety-five nondemented patients categorized as having normal cognition (n = 31), frontostriatal (n = 14), posterior cortical (n = 20), or mixed (n = 30) cognitive subtype had a 3 T resting-state functional magnetic resonance imaging scan. Twenty-four age-matched healthy controls (HCs) were also included. A group-level independent component analysis was performed to identify resting-state networks, and the selected components were subdivided into 564 cortical regions in addition to 26 basal ganglia regions. Global intra- and inter-network connectivity along with global and local efficiencies was compared between groups. The network-based statistics approach was used to identify connections significantly different between groups.

RESULTS

Patients with posterior cortical deficits had increased intra-network functional connectivity (FC) within the basal ganglia network compared with patients with frontostriatal deficits. Patients with frontostriatal deficits had reduced inter-network FC between several networks, including the visual, default-mode, sensorimotor, salience, dorsal attentional, basal ganglia, and frontoparietal networks, compared with HCs, patients with normal cognition, and patients with a posterior cortical subtype. Similar results were also found between patients with a mixed subtype and HCs.

CONCLUSION

MCI subtypes are associated with specific changes in resting-state FC. Longitudinal studies are needed to determine the predictive potential of these markers regarding the risk of developing dementia. © 2021 International Parkinson and Movement Disorder Society.

Detection of visual and frontoparietal network perfusion deficits in Parkinson's disease dementia

Mild cognitive impairment of Parkinson's disease (PD) may be an early manifestation that may progressively worsen to dementia. Cognitive decline has been associated with changes in the brain perfusion pattern. This study aimed to evaluate cerebral blood flow (CBF) deficits specific to different stages of cognitive decline. Seventeen patients with cognitively normal PD (PD-CN), 18 patients with PD with mild cognitive impairment (PD-MCI), and 16 patients with PD with dementia (PDD) were included in this study. The participants were scanned using a 3 T Philips MRI scanner. Arterial spin labelling magnetic resonance (ASL-MR) images were acquired, followed by calculation of the CBF maps, and registration onto the MNI152 brain atlas. A whole-brain voxel-based CBF comparison was performed among the patient groups using age as a covariate. The mean age of patients with PDD was significantly higher than that of patients with PD-MCI (P = 0.015) and PD-CN (P = 0.001). The CBF values of the three groups were significantly different in the left cuneus of the visual network (VN), left inferior frontal gyrus of the frontoparietal network (FPN), and left dorsomedial nucleus of the thalamus. PDD had lower perfusion values than PD-MCI group in the same regions detected in the main group analysis. Additionally, comparison of PDD with PD-CN and non-demented groups revealed that the perfusion reduction extended into the bilateral cuneus of the VN, bilateral thalami, and left inferior frontal gyrus of the FPN. PDD could be separated from PD-MCI and PD-CN stages with CBF deficits in non-dopaminergically mediated posterior and dopaminergically mediated frontal networks.

Topologically convergent and divergent morphological gray matter networks in early-stage Parkinson's disease with and without mild cognitive impairment

Patients with Parkinson's disease with mild cognitive impairment (PD‐M) progress to dementia more frequently than those with normal cognition (PD‐N), but the underlying neurobiology remains unclear. This study aimed to define the specific morphological brain network alterations in PD‐M, and explore their potential diagnostic value. Twenty‐four PD‐M patients, 17 PD‐N patients, and 29 healthy controls (HC) underwent a structural MRI scan. Similarity between interregional gray matter volume distributions was used to construct individual morphological brain networks. These were analyzed using graph theory and network‐based statistics (NBS), and their relationship to neuropsychological tests was assessed. Support vector machine (SVM) was used to perform individual classification. Globally, compared with HC, PD‐M showed increased local efficiency (p = .001) in their morphological networks, while PD‐N showed decreased normalized path length (p = .008). Locally, similar nodal deficits were found in the rectus and lingual gyrus, and cerebellum of both PD groups relative to HC; additionally in PD‐M nodal deficits involved several frontal and parietal regions, correlated with cognitive scores. NBS found that similar connections were involved in the default mode and cerebellar networks of both PD groups (to a greater extent in PD‐M), while PD‐M, but not PD‐N, showed altered connections involving the frontoparietal network. Using connections identified by NBS, SVM allowed discrimination with high accuracy between PD‐N and HC (90%), PD‐M and HC (85%), and between the two PD groups (65%). These results suggest that default mode and cerebellar disruption characterizes PD, more so in PD‐M, whereas frontoparietal disruption has diagnostic potential.

Posterior Cortical Cognitive Deficits Are Associated With Structural Brain Alterations in Mild Cognitive Impairment in Parkinson's Disease

Context: Cognitive impairments are common in patients with Parkinson's disease (PD) and are heterogeneous in their presentation. The "dual syndrome hypothesis" suggests the existence of two distinct subtypes of mild cognitive impairment (MCI) in PD: a frontostriatal subtype with predominant attentional and/or executive deficits and a posterior cortical subtype with predominant visuospatial, memory, and/or language deficits. The latter subtype has been associated with a higher risk of developing dementia. Objective: The objective of this study was to identify structural modifications in cortical and subcortical regions associated with each PD-MCI subtype. Methods: One-hundred and fourteen non-demented PD patients underwent a comprehensive neuropsychological assessment as well as a 3T magnetic resonance imaging scan. Patients were categorized as having no cognitive impairment (n = 41) or as having a frontostriatal (n = 16), posterior cortical (n = 25), or a mixed (n = 32) MCI subtype. Cortical regions were analyzed using a surface-based Cortical thickness (CTh) method. In addition, the volumes, shapes, and textures of the caudate nuclei, hippocampi, and thalami were studied. Tractometric analyses were performed on associative and commissural white matter (WM) tracts. Results: There were no between-group differences in volumetric measurements and cortical thickness. Shape analyses revealed more abundant and more extensive deformations fields in the caudate nuclei, hippocampi, and thalami in patients with posterior cortical deficits compared to patients with no cognitive impairment. Decreased fractional anisotropy (FA) and increased mean diffusivity (MD) were also observed in the superior longitudinal fascicle, the inferior fronto-occipital fascicle, the striato-parietal tract, and the anterior and posterior commissural tracts. Texture analyses showed a significant difference in the right hippocampus of patients with a mixed MCI subtype. Conclusion: PD-MCI patients with posterior cortical deficits have more abundant and more extensive structural alterations independently of age, disease duration, and severity, which may explain why they have an increased risk of dementia.

Neural signatures of predictive language processing in Parkinson's disease with and without mild cognitive impairment

Cognitive deficits are common in Parkinson's disease (PD), with some PD patients meeting criteria for mild cognitive impairment (MCI). An unaddressed question is whether linguistic prediction is preserved in PD. This ability is nowadays deemed crucial for achieving fast and efficient comprehension, and it may be negatively impacted by cognitive deterioration in PD. To fill this gap of knowledge, we used event-related potentials (ERPs) to evaluate mechanisms of linguistic prediction in a sample of PD patients (on dopamine compensation) with and without MCI. To this end, participants read sentence contexts that were predictive or not about a sentence-final word. The final word appeared after one sec, matching or mismatching the prediction. The introduction of the interval allowed to capture neural responses both before and after sentence-final words, reflecting semantic anticipation and semantic processing. PD patients with normal cognition (N = 58) showed ERP responses comparable to those of matched controls. Specifically, in predictive contexts, a slow negative potential developed prior to sentence-final words, reflecting semantic anticipation. Later, expected words elicited reduced N400 responses (compared to unexpected words), indicating facilitated semantic processing. PD patients with MCI (N = 20) showed, in addition, a prolongation of the N400 congruency effect (compared to matched PD patients without MCI), indicating that further cognitive decline impacts semantic processing. Finally, lower verbal fluency scores correlated with prolonged N400 congruency effects and with reduced pre-word differences in all PD patients (N = 78). This relevantly points to a role of deficits in temporal-dependent mechanisms in PD, besides prototypical frontal dysfunction, in altered semantic anticipation and semantic processing during sentence comprehension.

Functional Connectome in Parkinson's Disease and Parkinsonism

PURPOSE OF REVIEW

There has been an exponential growth in functional connectomics research in neurodegenerative disorders. This review summarizes the recent findings and limitations of the field in Parkinson's disease (PD) and atypical parkinsonian syndromes.

RECENT FINDINGS

Increasingly more sophisticated methods ranging from seed-based to network and whole-brain dynamic functional connectivity have been used. Results regarding the disruption in the functional connectome vary considerably based on disease severity and phenotypes, and treatment status in PD. Non-motor symptoms of PD also link to the dysfunction in heterogeneous networks. Studies in atypical parkinsonian syndromes are relatively scarce. An important clinical goal of functional connectomics in neurodegenerative disorders is to establish the presence of pathology, track disease progression, predict outcomes, and monitor treatment response. The obstacles of reliability and reproducibility in the field need to be addressed to improve the potential of the functional connectome as a biomarker for these purposes in PD and atypical parkinsonian syndromes.

Factor analysis-derived cognitive profile predicting early dementia conversion in PD

OBJECTIVES

To investigate which baseline neuropsychological profile predicts the risk of developing dementia in early-stage Parkinson disease (PD).

METHODS

We retrospectively reviewed detailed medical records of 350 drug-naive patients with early-stage PD (follow-up >3 years) who underwent a detailed neuropsychological test at initial assessment. Factor analysis was conducted to determine cognitive profiles that yielded 4 cognitive function factors: factor 1, visual memory/visuospatial; factor 2, verbal memory; factor 3, frontal/executive; and factor 4, attention/working memory/language. Subsequently, we assessed the effect of these cognitive function factors on the risk for dementia conversion. We also constructed a nomogram to calculate the risk for developing dementia over a 5-year follow-up period based on these cognitive profiles.

RESULTS

Cox regression analysis demonstrated that a higher composite score of factor 1 (hazard ratio [HR] 0.558, 95% confidence interval [CI] 0.427-0.730), factor 2 (HR 0.768, 95% CI 0.596-0.991), and factor 3 (HR 0.425, 95% CI 0.305-0.593) was associated with a lower risk for dementia conversion, while factor 3 had the most predictive power. The nomogram had a fair ability (Heagerty integrated area under the curve 0.763) to estimate the risk for dementia conversion within 5 years. The composite scores of factor 3 contributed more to the occurrence of dementia in PD than those of the other cognitive function factors.

CONCLUSIONS

These findings suggest that these factor analysis-derived cognitive profiles can be used to predict dementia conversion in early-stage PD. In addition, frontal/executive dysfunction contributes most to the occurrence of dementia in PD.

Common and unique connectivity at the interface of motor, neuropsychiatric, and cognitive symptoms in Parkinson's disease: A commonality analysis

Parkinson's disease (PD) is characterized by overlapping motor, neuropsychiatric, and cognitive symptoms. Worse performance in one domain is associated with worse performance in the other domains. Commonality analysis (CA) is a method of variance partitioning in multiple regression, used to separate the specific and common influence of collinear predictors. We apply, for the first time, CA to the functional connectome to investigate the unique and common neural connectivity underlying the interface of the symptom domains in 74 non-demented PD subjects. Edges were modeled as a function of global motor, cognitive, and neuropsychiatric scores. CA was performed, yielding measures of the unique and common contribution of the symptom domains. Bootstrap confidence intervals were used to determine the precision of the estimates and to directly compare each commonality coefficient. The overall model identified a network with the caudate nucleus as a hub. Neuropsychiatric impairment accounted for connectivity in the caudate-dorsal anterior cingulate and caudate-right dorsolateral prefrontal-right inferior parietal circuits, while caudate-medial prefrontal connectivity reflected a unique effect of both neuropsychiatric and cognitive impairment. Caudate-precuneus connectivity was explained by both unique and shared influence of neuropsychiatric and cognitive symptoms. Lastly, posterior cortical connectivity reflected an interplay of the unique and common effects of each symptom domain. We show that CA can determine the amount of variance in the connectome that is unique and shared amongst motor, neuropsychiatric, and cognitive symptoms in PD, thereby improving our ability to interpret the data while gaining novel insight into networks at the interface of these symptom domains.

Mild behavioral impairment in Parkinson's disease is associated with altered corticostriatal connectivity

•

Mild behavioral impairment in PD is linked to altered corticostriatal connectivity.

•

PD-MBI have less connectivity between the striatum and the DMN.

•

PD-MBI have increased atrophy of the SAN.

•

Caudate head and dorsal putamen connectivity is related to MBI-C scores in PD.

•

Caudate head-precuneus connectivity is linked to both MBI and MoCA scores.

Background

Mild behavioral impairment (MBI) is a syndrome characterized by later life onset, sustained neuropsychiatric symptoms as a marker of dementia risk. In Parkinson's disease (PD), MBI has been associated with worse cognitive abilities and increased cortical atrophy. However, the circuit level correlates of MBI have not been investigated in this population. Our objective was to investigate the relationship between MBI and corticostriatal connectivity in PD patients. This emphasis on corticostriatal connectivity was due to the significant role of these circuits in neuropsychiatric and cognitive symptoms across disease conditions.

Methods

Seventy-four non-demented patients with PD were administered the MBI-checklist, and classified as having high MBI (PD-MBI; n = 21) or low MBI scores (PD-noMBI; n = 53). Corticostriatal connectivity was assessed with both an atlas and seed-based analysis. The atlas analysis consisted of calculating the average connectivity between the striatal network and the default mode (DMN), central executive (CEN), and saliency networks (SAN). Structural measurements of cortical thickness and volume were also assessed. PD-MBI and PD-noMBI patients were compared, along with a group of age matched healthy control subjects (HC; n = 28). Subsequently, a seed analysis assessed the relationship of MBI scores with the connectivity of twelve seeds within the striatum while controlling for cognitive ability. A complementary analysis assessed the relationship between striatal connectivity and cognition, while controlling for MBI-C.

Results

PD-MBI demonstrated decreased connectivity between the striatum and both the DMN and SAN compared to PD-noMBI and HC. The decreased connectivity between the striatum and the SAN was explained partly by increased atrophy within the SAN in PD-MBI. The seed analysis revealed a relationship between higher MBI scores and lower connectivity of the left caudate head to the dorsal anterior cingulate cortex and left middle frontal gyrus. Higher MBI-C scores were also related to decreased connectivity of the right caudate head with the anterior cingulate cortex, precuneus, and left supramarginal gyrus, as well as increased connectivity to the left hippocampus and right cerebellar hemisphere. Caudate-precuneus connectivity was independently associated with both global behavioural and cognitive scores.

Conclusion

These results suggest PD-MBI is associated with altered corticostriatal connectivity, particularly between the head of the caudate and cortical regions associated with the DMN and SAN. In particular, caudate-precuneus connectivity is associated with both global behavioral and cognitive symptoms in PD.

Connectomics and molecular imaging in neurodegeneration

Our understanding on human neurodegenerative disease was previously limited to clinical data and inferences about the underlying pathology based on histopathological examination. Animal models and in vitro experiments have provided evidence for a cell-autonomous and a non-cell-autonomous mechanism for the accumulation of neuropathology. Combining modern neuroimaging tools to identify distinct neural networks (connectomics) with target-specific positron emission tomography (PET) tracers is an emerging and vibrant field of research with the potential to examine the contributions of cell-autonomous and non-cell-autonomous mechanisms to the spread of pathology. The evidence provided here suggests that both cell-autonomous and non-cell-autonomous processes relate to the observed in vivo characteristics of protein pathology and neurodegeneration across the disease spectrum. We propose a synergistic model of cell-autonomous and non-cell-autonomous accounts that integrates the most critical factors (i.e., protein strain, susceptible cell feature and connectome) contributing to the development of neuronal dysfunction and in turn produces the observed clinical phenotypes. We believe that a timely and longitudinal pursuit of such research programs will greatly advance our understanding of the complex mechanisms driving human neurodegenerative diseases.