Houston, We Have AI Problem! Quality Issues with Neuroimaging-Based Artificial Intelligence in Parkinson's Disease: A Systematic Review.

In recent years, many neuroimaging studies have applied artificial intelligence (AI) to facilitate existing challenges in Parkinson's disease (PD) diagnosis, prognosis, and intervention. The aim of this systematic review was to provide an overview of neuroimaging-based AI studies and to assess their methodological quality. A PubMed search yielded 810 studies, of which 244 that investigated the utility of neuroimaging-based AI for PD diagnosis, prognosis, or intervention were included. We systematically categorized studies by outcomes and rated them with respect to five minimal quality criteria (MQC) pertaining to data splitting, data leakage, model complexity, performance reporting, and indication of biological plausibility. We found that the majority of studies aimed to distinguish PD patients from healthy controls (54%) or atypical parkinsonian syndromes (25%), whereas prognostic or interventional studies were sparse. Only 20% of evaluated studies passed all five MQC, with data leakage, non-minimal model complexity, and reporting of biological plausibility as the primary factors for quality loss. Data leakage was associated with a significant inflation of accuracies. Very few studies employed external test sets (8%), where accuracy was significantly lower, and 19% of studies did not account for data imbalance. Adherence to MQC was low across all observed years and journal impact factors. This review outlines that AI has been applied to a wide variety of research questions pertaining to PD; however, the number of studies failing to pass the MQC is alarming. Therefore, we provide recommendations to enhance the interpretability, generalizability, and clinical utility of future AI applications using neuroimaging in PD. © 2024 The Author(s). Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.

Neurophysiological markers of motor compensatory mechanisms in early Parkinson's disease.

Compensatory mechanisms in Parkinson's disease are defined as the changes that the brain uses to adapt to neurodegeneration and progressive dopamine reduction. Motor compensation in early Parkinson's disease could, in part, be responsible for a unilateral onset of clinical motor signs despite the presence of bilateral nigrostriatal degeneration. Although several mechanisms have been proposed for compensatory adaptations in Parkinson's disease, the underlying pathophysiology is unclear. Here, we investigate motor compensation in Parkinson's disease by investigating the relationship between clinical signs, dopamine transporter imaging data and neurophysiological measures of the primary motor cortex (M1), using transcranial magnetic stimulation in presymptomatic and symptomatic hemispheres of patients. In this cross-sectional, multicentre study, we screened 82 individuals with Parkinson's disease. Patients were evaluated clinically in their medication OFF state using standardized scales. Sixteen Parkinson's disease patients with bilateral dopamine transporter deficit in the putamina but unilateral symptoms were included. Twenty-eight sex- and age-matched healthy controls were also investigated. In all participants, we tested cortical excitability using single- and paired-pulse techniques, interhemispheric inhibition and cortical plasticity with paired associative stimulation. Data were analysed with ANOVAs, multiple linear regression and logistic regression models. Individual coefficients of motor compensation were defined in patients based on clinical and imaging data, i.e. the motor compensation coefficient. The motor compensation coefficient includes an asymmetry score to balance motor and dopamine transporter data between the two hemispheres, in addition to a hemispheric ratio accounting for the relative mismatch between the magnitude of motor signs and dopaminergic deficit. In patients, corticospinal excitability and plasticity were higher in the presymptomatic compared with the symptomatic M1. Also, interhemispheric inhibition from the presymptomatic to the symptomatic M1 was reduced. Lower putamen binding was associated with higher plasticity and reduced interhemispheric inhibition in the presymptomatic hemisphere. The motor compensation coefficient distinguished the presymptomatic from the symptomatic hemisphere. Finally, in the presymptomatic hemisphere, a higher motor compensation coefficient was associated with lower corticospinal excitability and interhemispheric inhibition and with higher plasticity. In conclusion, the present study suggests that motor compensation involves M1-striatal networks and intercortical connections becoming more effective with progressive loss of dopaminergic terminals in the putamen. The balance between these motor networks seems to be driven by cortical plasticity.

Dynamic properties in functional connectivity changes and striatal dopamine deficiency in Parkinson's disease.

Recent studies in Parkinson's disease (PD) patients reported disruptions in dynamic functional connectivity (dFC, i.e., a characterization of spontaneous fluctuations in functional connectivity over time). Here, we assessed whether the integrity of striatal dopamine terminals directly modulates dFC metrics in two separate PD cohorts, indexing dopamine-related changes in large-scale brain network dynamics and its implications in clinical features. We pooled data from two disease-control cohorts reflecting early PD. From the Parkinson's Progression Marker Initiative (PPMI) cohort, resting-state functional magnetic resonance imaging (rsfMRI) and dopamine transporter (DaT) single-photon emission computed tomography (SPECT) were available for 63 PD patients and 16 age- and sex-matched healthy controls. From the clinical research group 219 (KFO) cohort, rsfMRI imaging was available for 52 PD patients and 17 age- and sex-matched healthy controls. A subset of 41 PD patients and 13 healthy control subjects additionally underwent 18F-DOPA-positron emission tomography (PET) imaging. The striatal synthesis capacity of 18F-DOPA PET and dopamine terminal quantity of DaT SPECT images were extracted for the putamen and the caudate. After rsfMRI pre-processing, an independent component analysis was performed on both cohorts simultaneously. Based on the derived components, an individual sliding window approach (44 s window) and a subsequent k-means clustering were conducted separately for each cohort to derive dFC states (reemerging intra- and interindividual connectivity patterns). From these states, we derived temporal metrics, such as average dwell time per state, state attendance, and number of transitions and compared them between groups and cohorts. Further, we correlated these with the respective measures for local dopaminergic impairment and clinical severity. The cohorts did not differ regarding age and sex. Between cohorts, PD groups differed regarding disease duration, education, cognitive scores and L-dopa equivalent daily dose. In both cohorts, the dFC analysis resulted in three distinct states, varying in connectivity patterns and strength. In the PPMI cohort, PD patients showed a lower state attendance for the globally integrated (GI) state and a lower number of transitions than controls. Significantly, worse motor scores (Unified Parkinson's Disease Rating Scale Part III) and dopaminergic impairment in the putamen and the caudate were associated with low average dwell time in the GI state and a low total number of transitions. These results were not observed in the KFO cohort: No group differences in dFC measures or associations between dFC variables and dopamine synthesis capacity were observed. Notably, worse motor performance was associated with a low number of bidirectional transitions between the GI and the lesser connected (LC) state across the PD groups of both cohorts. Hence, in early PD, relative preservation of motor performance may be linked to a more dynamic engagement of an interconnected brain state. Specifically, those large-scale network dynamics seem to relate to striatal dopamine availability. Notably, most of these results were obtained only for one cohort, suggesting that dFC is impacted by certain cohort features like educational level, or disease severity. As we could not pinpoint these features with the data at hand, we suspect that other, in our case untracked, demographical features drive connectivity dynamics in PD. PRACTITIONER POINTS: Exploring dopamine's role in brain network dynamics in two Parkinson's disease (PD) cohorts, we unraveled PD-specific changes in dynamic functional connectivity. Results in the Parkinson's Progression Marker Initiative (PPMI) and the KFO cohort suggest motor performance may be linked to a more dynamic engagement and disengagement of an interconnected brain state. Results only in the PPMI cohort suggest striatal dopamine availability influences large-scale network dynamics that are relevant in motor control.

Diagnostics and treatment of impulse control disorders, psychosis and delirium: systemic review-based recommendations - guideline "Parkinson's disease" of the German Society of Neurology.

BACKGROUND AND OBJECTIVE: Impulse control disorders (ICD), psychosis and delirium are part of the spectrum of behavioural changes associated with Parkinson's disease (PD). The diagnostic and therapeutic management of these rather complex neuropsychiatric conditions has been updated in the clinical guideline by the German Society of Neurology (DGN). METHODS: Recommendations are based on a systematic literature reviews, other relevant guidelines and expert opinion. RESULTS: Patients receiving dopamine agonists (DA) therapy should be informed about the symptoms and risks of an ICD and should be routinely screened for ICD symptoms. In the presence of an ICD, DA should be reduced or discontinued and psychotherapeutic treatment may be considered. Non-oral therapies (levodopa/carbidopa intestinal gel infusion or deep brain stimulation) may also be an option for appropriate candidates. Psychosis in PD often has a gradual onset. Cognitive and affective disorders, psychiatric and medical comorbidities as well as polypharmacy are risk factors for a psychosis. Non-pharmacological treatments should be implemented as soon as possible and anti-parkinsonian medications should be adjusted/reduced if feasible. For psychosis associated with PD, quetiapine or clozapine should be used on an as-needed basis and for as short a time as is necessary, with safety monitoring. Delirium in PD may be underdiagnosed due to an overlap with chronic neuropsychiatric features of PD. Although transient by definition, delirium in PD can lead to permanent cognitive decline, motor impairment and increased mortality. Management of delirium includes pharmacological and non-pharmacological interventions. CONCLUSION: The updated guideline encompasses the evidence-based diagnostic, non-pharmacological and pharmacological management of ICD, psychosis and delirium in PD.

Pharmacological Enhancement of Dopamine Neurotransmission Does Not Affect Illusory Pattern Perception.

Psychotic symptoms and delusional beliefs have been linked to dopamine transmission in both healthy and clinical samples and are assumed to result at least in part from perceiving illusory patterns in noise. However, the existing literature on the role of dopamine in detecting patterns in noise is inconclusive. To address this issue, we assessed the effect of manipulating dopaminergic neurotransmission on illusory pattern perception in healthy individuals (n = 48, n = 19 female) in a double-blind placebo-controlled within-subjects design (see preregistration at https://osf.io/a4k9j/). We predicted individuals on versus off ʟ-DOPA to be more likely to perceive illusory patterns, specifically objects in images containing only noise. Using a signal detection model, however, we found no credible evidence that ʟ-DOPA compared with placebo increased false alarm rates. Further, ʟ-DOPA did not reliably modulate measures of accuracy, discrimination sensitivity, and response bias. In all cases, Bayesian statistics revealed strong evidence in favor of the null hypothesis. The task design followed previous work on illusory pattern perception and comprised a limited number of items per condition. The results therefore need to be interpreted with caution, as power was limited. Future studies should address illusory pattern perception using more items and take into account potential dose-dependent effects and differential effects in healthy versus clinical samples.

Progressive Supranuclear Palsy: Subcortical Tau Depositions Are Associated with Cortical Perfusion in Frontal and Limbic Regions.

 In progressive supranuclear palsy (PSP), subcortical tau and cortical perfusion can be assessed using the tracer [18F]PI-2620. We investigated if subcortical tau (globus pallidus internus, dentate nucleus) and frontal/limbic perfusion correlate in a cohort of 32 PSP patients. Tau in subcortical regions showed significant negative correlation with perfusion in limbic cortex. Perfusion in frontal regions was negatively associated with tau in both subcortical regions, but the significance threshold was only passed for the dentate nucleus. A reason could be a diaschisis-like phenomenon; that is, subcortical tau could lead to reduced connectivity to frontal regions and, thereby, to decreased perfusion.

In a study of 32 patients with progressive supranuclear palsy (PSP), we used a molecular imaging tracer called [18F]PI-2620 to measure two things: the presence of a protein called tau in deep brain areas (specifically, the globus pallidus internus and dentate nucleus) and the function of the brain's cortex by assessing blood flow (perfusion). We found that higher amounts of tau in these deep brain areas were associated with reduced blood flow in the limbic cortex, which is involved in emotion regulation. Also, the frontal areas of the brain showed reduced blood flow related to tau in these deep brain regions. However, this connection was statistically significant only for the dentate nucleus. This study suggests that the buildup of tau protein in deeper brain areas can disrupt function in parts of the brain's cortex, highlighting the damaging role of tau in PSP.

Sharing brain imaging data in the Open Science era: how and why?

The sharing of human neuroimaging data has great potential to accelerate the development of imaging biomarkers in neurological and psychiatric disorders; however, major obstacles remain in terms of how and why to share data in the Open Science context. In this Health Policy by the European Cluster for Imaging Biomarkers, we outline the current main opportunities and challenges based on the results of an online survey disseminated among senior scientists in the field. Although the scientific community fully recognises the importance of data sharing, technical, legal, and motivational aspects often prevent active adoption. Therefore, we provide practical advice on how to overcome the technical barriers. We also call for a harmonised application of the General Data Protection Regulation across EU countries. Finally, we suggest the development of a system that makes data count by recognising the generation and sharing of data as a highly valuable contribution to the community.

Putaminal dopamine modulates movement motivation in Parkinson's disease.

The relative inability to produce effortful movements is the most specific motor sign of Parkinson's disease, which is primarily characterized by loss of dopaminergic terminals in the putamen. The motor motivation hypothesis suggests that this motor deficit may not reflect a deficiency in motor control per se, but a deficiency in cost-benefit considerations for motor effort. For the first time, we investigated the quantitative effect of dopamine depletion on the motivation of motor effort in Parkinson's disease. A total of 21 early-stage, unmedicated patients with Parkinson's disease and 26 healthy controls were included. An incentivized force task was used to capture the amount of effort participants were willing to invest for different monetary incentive levels and dopamine transporter depletion in the bilateral putamen was assessed. Our results demonstrate that patients with Parkinson's disease applied significantly less grip force than healthy controls, especially for low incentive levels. Congruously, decrease of motor effort with greater loss of putaminal dopaminergic terminals was most pronounced for low incentive levels. This signifies that putaminal dopamine is most critical to motor effort when the trade-off with the benefit is poor. Taken together, we provide direct evidence that the reduction of effortful movements in Parkinson's disease depends on motivation and that this effect is associated with putaminal dopaminergic degeneration.

Structural underpinnings and long-term effects of resilience in Parkinson's disease.

Resilience in neuroscience generally refers to an individual's capacity to counteract the adverse effects of a neuropathological condition. While resilience mechanisms in Alzheimer's disease are well-investigated, knowledge regarding its quantification, neurobiological underpinnings, network adaptations, and long-term effects in Parkinson's disease is limited. Our study involved 151 Parkinson's patients from the Parkinson's Progression Marker Initiative Database with available Magnetic Resonance Imaging, Dopamine Transporter Single-Photon Emission Computed Tomography scans, and clinical information. We used an improved prediction model linking neuropathology to symptom severity to estimate individual resilience levels. Higher resilience levels were associated with a more active lifestyle, increased grey matter volume in motor-associated regions, a distinct structural connectivity network and maintenance of relative motor functioning for up to a decade. Overall, the results indicate that relative maintenance of motor function in Parkinson's patients may be associated with greater neuronal substrate, allowing higher tolerance against neurodegenerative processes through dynamic network restructuring.

Improved Tau PET SUVR Quantification in 4-Repeat Tau Phenotypes with [18F]PI-2620.

We used a new data-driven methodology to identify a set of reference regions that enhanced the quantification of the SUV ratio of the second-generation tau tracer 2-(2-([18F]fluoro)pyridin-4-yl)-9H-pyrrolo[2,3-b:4,5-c']dipyridine ([18F]PI-2620) in a group of patients clinically diagnosed with 4-repeat tauopathy, specifically progressive supranuclear palsy or cortical basal syndrome. The study found that SUV ratios calculated using the identified reference regions (i.e., fusiform gyrus and crus-cerebellum) were significantly associated with symptom severity and disease duration. This establishes, for the first time to our knowledge, the suitability of [18F]PI-2620 for tracking disease progression in this 4-repeat disease population. This is an important step toward increased clinical utility, such as patient stratification and monitoring in disease-modifying treatment trials. Additionally, the applied methodology successfully optimized reference regions for automated detection of brain imaging tracers. This approach may also hold value for other brain imaging tracers.