Imaging in Parkinson's Disease

Mining imaging and clinical data with machine learning approaches for the diagnosis and early detection of Parkinson's disease

Parkinson’s disease (PD) is a common, progressive, and currently incurable neurodegenerative movement disorder. The diagnosis of PD is challenging, especially in the differential diagnosis of parkinsonism and in early PD detection. Due to the advantages of machine learning such as learning complex data patterns and making inferences for individuals, machine-learning techniques have been increasingly applied to the diagnosis of PD, and have shown some promising results. Machine-learning-based imaging applications have made it possible to help differentiate parkinsonism and detect PD at early stages automatically in a number of neuroimaging studies. Comparative studies have shown that machine-learning-based SPECT image analysis applications in PD have outperformed conventional semi-quantitative analysis in detecting PD-associated dopaminergic degeneration, performed comparably well as experts’ visual inspection, and helped improve PD diagnostic accuracy of radiologists. Using combined multi-modal (imaging and clinical) data in these applications may further enhance PD diagnosis and early detection. To integrate machine-learning-based diagnostic applications into clinical systems, further validation and optimization of these applications are needed to make them accurate and reliable. It is anticipated that machine-learning techniques will further help improve differential diagnosis of parkinsonism and early detection of PD, which may reduce the error rate of PD diagnosis and help detect PD at pre-motor stage to make it possible for early treatments (e.g., neuroprotective treatment) to slow down PD progression, prevent severe motor symptoms from emerging, and relieve patients from suffering.

Experimental Models of Cognitive Impairment for Use in Parkinson's Disease Research: The Distance Between Reality and Ideal

Parkinson’s disease (PD) is the second most common neurodegenerative disease. Cognitive impairment is one of the key non-motor symptoms of PD, affecting both mortality and quality of life. However, there are few experimental studies on the pathology and treatments of PD with mild cognitive impairment (PD-MCI) and PD dementia (PDD) due to the lack of representative models. To identify new strategies for developing representative models, we systematically summarized previous studies on PD-MCI and PDD and compared differences between existing models and diseases. Our initial search identified 5432 articles, of which 738 were duplicates. A total of 227 articles met our inclusion criteria and were included in the analysis. Models fell into three categories based on model design: neurotoxin-induced, transgenic, and combined. Although the neurotoxin-induced experimental model was the most common type that was used during every time period, transgenic and combined experimental models have gained significant recent attention. Unfortunately, there remains a big gap between ideal and actual experimental models. While each model has its own disadvantages, there have been tremendous advances in the development of PD models of cognitive impairment, and almost every model can verify a hypothesis about PD-MCI or PDD. Finally, our proposed strategies for developing novel models are as follows: a set of plans that integrate symptoms, biochemistry, neuroimaging, and other objective indicators to judge and identify that the novel model plays a key role in new strategies for developing representative models; novel models should simulate different clinical features of PD-MCI or PDD; inducible α-Syn overexpression and SH-SY5Y-A53T cellular models are good candidate models of PD-MCI or PDD.

Different Alterations of Agonist and Antagonist Binding to 5-HT1A Receptor in a Rat Model of Parkinson’s Disease and Levodopa-Induced Dyskinesia: A MicroPET Study

Background: The gold-standard treatment for Parkinson’s disease is L-DOPA, which in the long term often leads to levodopa-induced dyskinesia. Serotonergic neurons are partially responsible for this, by converting L-DOPA into dopamine leading to its uncontrolled release as a “false neurotransmitter”. The stimulation of 5-HT1A receptors can reduce involuntary movements but this mechanism is poorly understood. Objective: This study aimed to investigate the functionality of 5-HT1A receptors using positron emission tomography in hemiparkinsonian rats with or without dyskinesia induced by 3-weeks daily treatment with L-DOPA. Imaging sessions were performed “off” L-DOPA. Methods: Each rat underwent a positron emission tomography scan with [18F]F13640, a 5-HT1AR agonist which labels receptors in a high affinity state for agonists, or with [18F]MPPF, a 5-HT1AR antagonist which labels all the receptors. Results: There were decreases of [18F]MPPF binding in hemiparkinsonian rats in cortical areas. In dyskinetic animals, changes were slighter but also found in other regions. In hemiparkinsonian rats, [18F]F13640 uptake was decreased bilaterally in the globus pallidus and thalamus. On the non-lesioned side, binding was increased in the insula, the hippocampus and the amygdala. In dyskinetic animals, [18F]F13640 binding was strongly increased in cortical and limbic areas, especially in the non-lesioned side. Conclusion: These data suggest that agonist and antagonist 5-HT1A receptor-binding sites are differently modified in Parkinson’s disease and levodopa-induced dyskinesia. In particular, these observations suggest a substantial involvement of the functional state of 5-HT1AR in levodopa-induced dyskinesia and emphasize the need to characterize this state using agonist radiotracers in physiological and pathological conditions.

Deep Phenotyping of Parkinson's Disease

Phenotype is the set of observable traits of an organism or condition. While advances in genetics, imaging, and molecular biology have improved our understanding of the underlying biology of Parkinson's disease (PD), clinical phenotyping of PD still relies primarily on history and physical examination. These subjective, episodic, categorical assessments are valuable for diagnosis and care but have left gaps in our understanding of the PD phenotype. Sensors can provide objective, continuous, real-world data about the PD clinical phenotype, increase our knowledge of its pathology, enhance evaluation of therapies, and ultimately, improve patient care. In this paper, we explore the concept of deep phenotyping-the comprehensive assessment of a condition using multiple clinical, biological, genetic, imaging, and sensor-based tools-for PD. We discuss the rationale for, outline current approaches to, identify benefits and limitations of, and consider future directions for deep clinical phenotyping.

Use of Overlapping Group LASSO Sparse Deep Belief Network to Discriminate Parkinson's Disease and Normal Control

As a medical imaging technology which can show the metabolism of the brain, 18F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) is of great value for the diagnosis of Parkinson's Disease (PD). With the development of pattern recognition technology, analysis of brain images using deep learning are becoming more and more popular. However, existing computer-aided-diagnosis technologies often over fit and have poor generalizability. Therefore, we aimed to improve a framework based on Group Lasso Sparse Deep Belief Network (GLS-DBN) for discriminating PD and normal control (NC) subjects based on FDG-PET imaging. In this study, 225 NC and 125 PD cohorts from Huashan and Wuxi 904 hospitals were selected. They were divided into the training & validation dataset and 2 test datasets. First, in the training & validation set, subjects were randomly partitioned 80:20, with multiple training iterations for the deep learning model. Next, Locally Linear Embedding was used as a dimension reduction algorithm. Then, GLS-DBN was used for feature learning and classification. Different sparse DBN models were used to compare datasets to evaluate the effectiveness of our framework. Accuracy, sensitivity, and specificity were examined to validate the results. Output variables of the network were also correlated with longitudinal changes of rating scales about movement disorders (UPDRS, H&Y). As a result, accuracy of prediction (90% in Test 1, 86% in Test 2) for classification of PD and NC patients outperformed conventional approaches. Output scores of the network were strongly correlated with UPDRS and H&Y (R = 0.705, p < 0.001; R = 0.697, p < 0.001 in Test 1; R = 0.592, p = 0.0018, R = 0.528, p = 0.0067 in Test 2). These results show the GLS-DBN is feasible method for early diagnosis of PD.

Dementia spectrum disorders: lessons learnt from decades with PET research

The dementia spectrum encompasses a range of disorders with complex diagnosis, pathophysiology and limited treatment options. Positron emission tomography (PET) imaging provides insights into specific neurodegenerative processes underlying dementia disorders in vivo. Here we focus on some of the most common dementias: Alzheimer's disease, Parkinsonism dementias including Parkinson's disease with dementia, dementia with Lewy bodies, progressive supranuclear palsy and corticobasal syndrome, and frontotemporal lobe degeneration. PET tracers have been developed to target specific proteinopathies (amyloid, tau and α-synuclein), glucose metabolism, cholinergic system and neuroinflammation. Studies have shown distinct imaging abnormalities can be detected early, in some cases prior to symptom onset, allowing disease progression to be monitored and providing the potential to predict symptom onset. Furthermore, advances in PET imaging have identified potential therapeutic targets and novel methods to accurately discriminate between different types of dementias in vivo. There are promising imaging markers with a clinical application on the horizon, however, further studies are required before they can be implantation into clinical practice.

Hybrid PET-MRI Applications in Movement Disorders

Even before the success of combined positron emission tomography and computed tomography (PET/CT), the neuroimaging community was conceiving the idea to integrate the positron emission tomography (PET), with very high molecular quantitative data but low spatial resolution, and magnetic resonance imaging (MRI), with high spatial resolution. Several technical limitations have delayed the use of a hybrid scanner in neuroimaging studies, including the full integration of the PET detector ring within the MRI system, the optimization of data acquisition, and the implementation of reliable methods for PET attenuation, motion correction, and joint image reconstruction. To be valid and useful in clinical and research settings, this instrument should be able to simultaneously acquire PET and MRI, and generate quantitative parametric PET images comparable to PET-CT. While post hoc co-registration of combined PET and MRI data acquired separately became the most reliable technique for the generation of "fused" PET-MRI images, only hybrid PET-MRI approach allows merging these measurements naturally and correlating them in a temporal manner. Furthermore, hybrid PET-MRI represents the most accurate tool to investigate in vivo the interplay between molecular and functional aspects of brain pathophysiology. Hybrid PET-MRI technology is still in the early stages in the movement disorders field, due to the limited availability of scanners with integrated optimized methodological models. This technology is ideally suited to investigate interactions between resting-state functional/arterial spin labeling MRI and [¹⁸F]FDG PET glucose metabolism in the evaluation of the brain "hubs" particularly vulnerable to neurodegeneration, areas with a high degree of connectivity and associated with an efficient synaptic neurotransmission. In Parkinson's disease, hybrid PET-MRI is also the ideal instrument to deeper explore the relationship between resting-state functional MRI and dopamine release at [¹¹C]raclopride PET challenge, in the identification of early drug-naïve Parkinson's disease patients at higher risk of motor complications and in the evaluation of the efficacy of novel neuroprotective treatment able to restore at the same time the altered resting state and the release of dopamine. In this chapter, we discuss the key methodological aspects of hybrid PET-MRI; the evidence in movement disorders of the key resting-state functional and perfusion MRI; [¹⁸F]FDG PET and [¹¹C]raclopride PET challenge studies; the potential advantages of using hybrid PET-MRI to investigate the pathophysiology of movement disorders and neurodegenerative diseases. Future directions of hybrid PET-MRI will be discussed alongside with up-to-date technological innovations on hybrid systems.

Imaging Transplantation in Movement Disorders

Cell replacement therapy with graft transplantation has been tested as a disease-modifying treatment in neurodegenerative diseases characterized by the damage of a predominant cell type, such as substantia nigra dopaminergic neurons in Parkinson's disease (PD) or striatal medium spiny projection neurons in Huntington's disease (HD). The results of these trials are mixed with success in preclinical and pilot open-label trials, which were not consistently reproduced in randomized controlled trials. Positron emission tomography (PET) and single photon emission computed tomography (SPECT) molecular imaging and functional magnetic resonance imaging allow the graft survival, and its relationship with the host tissues to be studied in vivo. In PD, PET with [¹⁸F]DOPA showed that graft survival does not necessarily correlate with the clinical improvement and PD patients with worse outcome had lower binding in the ventral striatum and a high serotonin ([¹¹C]DASB PET) to dopamine ([¹⁸F]DOPA PET) ratio in the grafted neurons. In HD, PET with [¹¹C]PK11195 showed the graft survival and the clinical responses may be related to the reactive activation of the host inflammatory/immune system. Findings from these studies have been used to refine study protocols and patient selection in current clinical trials, which includes identifying suitable candidates for transplantation using imaging markers and employing multiple and/or novel PET tracers to better assess graft functions and inflammatory responses to grafts.

PET Imaging in Movement Disorders

Positron emission tomography (PET) has revealed key insights into the pathophysiology of movement disorders. This paper will focus on how PET investigations of pathophysiology are particularly relevant to Parkinson disease, a neurodegenerative condition usually starting later in life marked by a varying combination of motor and nonmotor deficits. Various molecular imaging modalities help to determine what changes in brain herald the onset of pathology; can these changes be used to identify presymptomatic individuals who may be appropriate for to-be-developed treatments that may forestall onset of symptoms or slow disease progression; can PET act as a biomarker of disease progression; can molecular imaging help enrich homogenous cohorts for clinical studies; and what other pathophysiologic mechanisms relate to nonmotor manifestations. PET methods include measurements of regional cerebral glucose metabolism and blood flow, selected receptors, specific neurotransmitter systems, postsynaptic signal transducers, and abnormal protein deposition. We will review each of these methodologies and how they are relevant to important clinical issues pertaining to Parkinson disease.