An entropy-based model for basal ganglia dysfunctions in movement disorders

Why so slow? Models of parkinsonian bradykinesia

Bradykinesia, or slowness of movement, is a defining feature of Parkinson disease (PD) and a major contributor to the negative effects on quality of life associated with this disorder and related conditions. A dominant pathophysiological model of bradykinesia in PD has existed for approximately 30 years and has been the basis for the development of several therapeutic interventions, but accumulating evidence has made this model increasingly untenable. Although more recent models have been proposed, they also appear to be flawed. In this Perspective, I consider the leading prior models of bradykinesia in PD and argue that a more functionally related model is required, one that considers changes that disrupt the fundamental process of accurate information transmission. In doing so, I review emerging evidence of network level functional connectivity changes, information transfer dysfunction and potential motor code transmission error and present a novel model of bradykinesia in PD that incorporates this evidence. I hope that this model may reconcile inconsistencies in its predecessors and encourage further development of therapeutic interventions.

Head movement dynamics in dystonia: a multi-centre retrospective study using visual perceptive deep learning

Dystonia is a neurological movement disorder characterised by abnormal involuntary movements and postures, particularly affecting the head and neck. However, current clinical assessment methods for dystonia rely on simplified rating scales which lack the ability to capture the intricate spatiotemporal features of dystonic phenomena, hindering clinical management and limiting understanding of the underlying neurobiology. To address this, we developed a visual perceptive deep learning framework that utilizes standard clinical videos to comprehensively evaluate and quantify disease states and the impact of therapeutic interventions, specifically deep brain stimulation. This framework overcomes the limitations of traditional rating scales and offers an efficient and accurate method that is rater-independent for evaluating and monitoring dystonia patients. To evaluate the framework, we leveraged semi-standardized clinical video data collected in three retrospective, longitudinal cohort studies across seven academic centres. We extracted static head angle excursions for clinical validation and derived kinematic variables reflecting naturalistic head dynamics to predict dystonia severity, subtype, and neuromodulation effects. The framework was also applied to a fully independent cohort of generalised dystonia patients for comparison between dystonia sub-types. Computer vision-derived measurements of head angle excursions showed a strong correlation with clinically assigned scores. Across comparisons, we identified consistent kinematic features from full video assessments encoding information critical to disease severity, subtype, and effects of neural circuit interventions, independent of static head angle deviations used in scoring. Our visual perceptive machine learning framework reveals kinematic pathosignatures of dystonia, potentially augmenting clinical management, facilitating scientific translation, and informing personalized precision neurology approaches.

Entropy and fractal analysis of brain-related neurophysiological signals in Alzheimer's and Parkinson's disease

Brain-related neuronal recordings, such as local field potential, electroencephalogram and magnetoencephalogram, offer the opportunity to study the complexity of the human brain at different spatial and temporal scales. The complex properties of neuronal signals are intrinsically related to the concept of 'scale-free' behavior and irregular dynamic, which cannot be fully described through standard linear methods, but can be measured by nonlinear indexes. A remarkable application of these analysis methods on electrophysiological recordings is the deep comprehension of the pathophysiology of neurodegenerative diseases, that has been shown to be associated to changes in brain activity complexity. In particular, a decrease of global complexity has been associated to Alzheimer's disease, while a local increase of brain signals complexity characterizes Parkinson's disease. Despite the recent proliferation of studies using fractal and entropy-based analysis, the application of these techniques is still far from clinical practice, due to the lack of an agreement about their correct estimation and a conclusive and shared interpretation. Along with the aim of helping towards the realization of a multidisciplinary audience to approach nonlinear methods based on the concepts of fractality and irregularity, this survey describes the implementation and proper employment of the mostly known and applied indexes in the context of Alzheimer's and Parkinson's diseases.

Basal ganglia functional connectivity network analysis does not support the 'noisy signal' hypothesis of Parkinson's disease

The 'noisy signal' hypothesis of basal ganglia dysfunction in Parkinson's disease (PD) suggests that major motor symptoms of the disorder are caused by the development of abnormal basal ganglia activity patterns resulting in the propagation of 'noisy' signals to target systems. While such abnormal activity patterns might be useful biomarkers for the development of therapeutic interventions, correlation between specific changes in activity and PD symptoms has been inconsistently demonstrated, and raises questions concerning the accuracy of the hypothesis. Here, we tested this hypothesis by considering three nodes of the basal ganglia network, the subthalamus, globus pallidus interna, and cortex during self-paced and cued movements in patients with PD. Interactions between these regions were analyzed using measures that assess both linear and non-linear relationships. Marked changes in the network are observed with dopamine state. Specifically, we detected functional disconnection of the basal ganglia from the cortex and higher network variability in untreated PD, but various patterns of directed functional connectivity with lower network variability in treated PD. When we examine the system output, significant correlation is observed between variability in the cortico-basal ganglia network and muscle activity variability but only in the treated state. Rather than supporting a role of the basal ganglia in the transmission of noisy signals in patients with PD, these findings suggest that cortico-basal ganglia network interactions by fault or design, in the treated Parkinsonian state, are actually associated with improved cortical network output variability.

Spectrum dependency to Rate and Spike Timing in neuronal spike trains

BACKGROUND

Spike trains are series of interspike intervals in a specific order that can be characterized by their probability distributions and order in time which refer to the concepts of rate and spike timing features. Periodic structure in the spike train can be reflected in oscillatory activities. Thus, there is a direct link between oscillator activities and the spike train. The proposed methods are to investigate the dependency of emerging oscillatory activities to the rate and the spike timing features.

METHOD

First, the circular statistics methods were compared to Fast Fourier Transform for best estimation of spectra. Second, two statistical tests were introduced to help make decisions regarding the dependency of spectrum, or individual frequencies, onto rate and spike timing. Third, the methodology is applied to in-vivo recordings of basal ganglia neurons in mouse, primate, and human. Finally, this novel framework is shown to allow the investigation of subsets of spikes contributing to individual oscillators.

RESULTS

Use of circular statistical methods, in comparison to FFT, minimizes spectral leakage. Using virtual spike trains, the Rate versus Timing Dependency Spectrum Test (or RTDs-Test) permits identifying spectral spike trains solely dependent on the rate feature from those that are also dependent on the spike timing feature. Similarly, the Rate versus Timing Dependency Frequency Test (or RTDf-Test), allows to identify individual oscillators with partial dependency on spike timing. Dependency on spike timing was found for all in-vivo recordings but only in few frequencies. The mapping in frequency and time of dependencies showed a dynamical process that may be organizing the basal ganglia function.

CONCLUSIONS

The methodology may improve our understanding of the emergence of oscillatory activities and, possibly, the relation between oscillatory activities and circuitry functions.

Challenges for future theories of Parkinson pathophysiology

Current theories on the basal ganglia-thalamic-cortical circuitry address the phenomena of hypokinesia and hyperkinesia. In this Perspective, we question whether the current models can address the orchestration of the motor units which is the common final pathway of the motor system. We conclude that the current theories do not to address this orchestration in health and disease. One alternative approach worthy of consideration is nonmonotonic nonlinear dynamics that contrast with a fundamentally linear or monotonic nonlinear approach that are presumed by current theories of basal ganglia-thalamic-cortical system. The purpose here is to make the case that current theories do presuppose a linear or monotonic nonlinear perspective which will be demonstrated as failing to adequately explicate the complex orchestration of motor unit activities in normal movement and in movement disorders. The notion of nonlinear dynamics is not new to neurophysiology; however, it is argued that it is new to the concepts of the physiology and pathophysiology of the basal ganglia-thalamic-cortical system. Providing a wholesale reconceptualization of the basal ganglia-thalamic-cortical system is beyond the scope of this effort. Rather, the contribution of the essay is convincing that there is a need to reconceptualize theories as nonlinear dynamical systems and there are metaphors and analogies from nonlinear science that can be productive in the reconsideration.

The Thermodynamic Consequences of Parkinson's Disease

Several lines of evidence point to a pervasive disturbance of energy balance in Parkinson's disease (PD). Weight loss, common and multifactorial, is the most observable sign of this. Bradykinesia may be best understood as an underinvestment of energy in voluntary movement. This accords with rodent experiments that emphasise the importance of dopamine in allocating motor energy expenditure. Oxygen consumption studies in PD suggest that, when activities are standardised for work performed, these inappropriate energy thrift settings are actually wasteful. That the dopaminergic deficit of PD creates a problem with energy efficiency highlights the role played by the basal ganglia, and by dopamine, in thermodynamic governance. This involves more than balancing energy, since living things maintain their internal order by controlling transformations of energy, resisting probabilistic trends to more random states. This review will also look at recent research in PD on the analysis of entropy-an information theory metric of predictability in a message-in recordings from the basal ganglia. Close relationships between energy and information converge around the concept of entropy. This is especially relevant to the motor system, which regulates energy exchange with the outside world through its flow of information. The malignant syndrome in PD, a counterpart of neuroleptic malignant syndrome, demonstrates how much thermodynamic disruption can result from breakdown of motor signalling in an extreme hypodopaminergic state. The macroenergetic disturbances of PD are consistent with a unifying hypothesis of dopamine's neurotransmitter actions-to adapt energy expenditure to prevailing economic circumstances.

Local field potential dynamics in the primate cortex in relation to parkinsonism reveled by machine learning: A comparison between the primary motor cortex and the supplementary area

The present study compares the cortical local field potentials (LFPs) in the primary motor cortex (M1) and the supplementary motor area (SMA) of non-human primates rendered Parkinsonian with administration of dopaminergic neurotoxin, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine. The dynamic of the LFPs was investigated under several mathematical frameworks and machine learning was used to discriminate the recordings based on these features between healthy, parkinsonian with off-medication and parkinsonian with on-medication states. The importance of each feature in the discrimination process was further investigated. The dynamic of the LFPs in M1 and SMA was affected regarding its variability (time domain analysis), oscillatory activities (frequency domain analysis) and complex patterns (non-linear domain analysis). Machine learning algorithms achieved accuracy near 0.90 for comparisons between conditions. The TreeBagger algorithm provided best accuracy. The relative importance of these features differed with the cortical location, condition and treatment. Overall, the most important features included beta oscillation, fractal dimension, gamma oscillation, entropy and asymmetry of amplitude fluctuation. The importance of features in discriminating between normal and pathological states, and on- or off-medication states depends on the pair-comparison and it is region-specific. These findings are discussed regarding the refinement of current models for movement disorders and the development of on-demand therapies.

Pallidal deep brain stimulation modulates excessive cortical high β phase amplitude coupling in Parkinson disease

OBJECTIVE

Deep brain stimulation (DBS) of the subthalamic nucleus (STN) and globus pallidus internus (GPi) are equally efficacious in the management of Parkinson disease (PD). Studies of STN-DBS have revealed a therapeutic reduction in excessive cortical β-γ phase-amplitude coupling (PAC). It is unclear whether this is specific to STN-DBS and potentially mediated by modulation of the hyperdirect pathway or if it is a generalizable mechanism seen with DBS of other targets. Moreover, it remains unclear how cortical signals are differentially modulated by movement versus therapy. To clarify, the effects of GPi-DBS and movement on cortical β power and β-γ PAC were examined.

METHODS

Right sensorimotor electrocorticographic signals were recorded in 10 PD patients undergoing GPi-DBS implantation surgery. We evaluated cortical β power and β-γ PAC during blocks of rest and contralateral hand movement (finger tapping) with GPi-DBS off and on.

RESULTS

Movement suppressed cortical low β power (P = 0.008) and high β-γ PAC (P = 0.028). Linear mixed effect modeling (LMEM) showed that power in low and high β bands are differentially modulated by movement (P = 0.022). GPi-DBS also results in a significant suppression of high β-γ PAC but without power modulation in either β sub-band (P = 0.008). Cortical high β-γ PAC is significantly correlated with severity of bradykinesia (Rho = 0.59, P = 0.045) and changes proportionally with therapeutic improvement (Rho = 0.61, P = 0.04).

CONCLUSIONS

Similar to STN-DBS, GPi-DBS reduces motor cortical β-γ PAC, like that also reported with dopaminergic mediations, suggesting it is a generalizable symptom biomarker in PD, independent of therapeutic target or proximity to the hyperdirect pathway.

Subthalamic neural entropy is a feature of freezing of gait in freely moving people with Parkinson's disease

The goal of this study was to investigate subthalamic (STN) neural features of Freezers and Non-Freezers with Parkinson's disease (PD), while freely walking without freezing of gait (FOG) and during periods of FOG, which were better elicited during a novel turning and barrier gait task than during forward walking.

METHODS

Synchronous STN local field potentials (LFPs), shank angular velocities, and ground reaction forces were measured in fourteen PD subjects (eight Freezers) off medication, OFF deep brain stimulation (DBS), using an investigative, implanted, sensing neurostimulator (Activa® PC+S, Medtronic, Inc.). Tasks included standing still, instrumented forward walking, stepping in place on dual forceplates, and instrumented walking through a turning and barrier course.

RESULTS

During locomotion without FOG, Freezers showed lower beta (13-30Hz) power (P=0.036) and greater beta Sample Entropy (P=0.032), than Non-Freezers, as well as greater gait asymmetry and arrhythmicity (P<0.05 for both). No differences in alpha/beta power and/or entropy were evident at rest. During periods of FOG, Freezers showed greater alpha (8-12Hz) Sample Entropy (P<0.001) than during walking without FOG.

CONCLUSIONS

A novel turning and barrier course was superior to FW in eliciting FOG. Greater unpredictability in subthalamic beta rhythms was evident during stepping without freezing episodes in Freezers compared to Non-Freezers, whereas greater unpredictability in alpha rhythms was evident in Freezers during FOG. Non-linear analysis of dynamic neural signals during gait in freely moving people with PD may yield greater insight into the pathophysiology of FOG; whether the increases in STN entropy are causative or compensatory remains to be determined. Some beta LFP power may be useful for rhythmic, symmetric gait and DBS parameters, which completely attenuate STN beta power may worsen rather than improve FOG.

Parkinsonian Balance Deficits Quantified Using a Game Industry Board and a Specific Battery of Four Paradigms

This study describes a cost-effective screening protocol for parkinsonism based on combined objective and subjective monitoring of balance function. Objective evaluation of balance function was performed using a game industry balance board and an automated analyses of the dynamic of the center of pressure in time, frequency, and non-linear domains collected during short series of stand up tests with different modalities and severity of sensorial deprivation. The subjective measurement of balance function was performed using the Dizziness Handicap Inventory questionnaire. Principal component analyses on both objective and subjective measurements of balance function allowed to obtained a specificity and selectivity for parkinsonian patients (vs. healthy subjects) of 0.67 and 0.71 respectively. The findings are discussed regarding the relevance of cost-effective balance-based screening system as strategy to meet the needs of broader and earlier screening for parkinsonism in communities with limited access to healthcare.

Neuronal Entropy-Rate Feature of Entopeduncular Nucleus in Rat Model of Parkinson's Disease

The function of the nigro-striatal pathway on neuronal entropy in the basal ganglia (BG) output nucleus, i.e. the entopeduncular nucleus (EPN) was investigated in the unilaterally 6-hyroxydopamine (6-OHDA)-lesioned rat model of Parkinson's disease (PD). In both control subjects and subjects with 6-OHDA lesion of dopamine (DA) the nigro-striatal pathway, a histological hallmark for parkinsonism, neuronal entropy in EPN was maximal in neurons with firing rates ranging between 15 and 25 Hz. In 6-OHDA lesioned rats, neuronal entropy in the EPN was specifically higher in neurons with firing rates above 25 Hz. Our data establishes that the nigro-striatal pathway controls neuronal entropy in motor circuitry and that the parkinsonian condition is associated with abnormal relationship between firing rate and neuronal entropy in BG output nuclei. The neuronal firing rates and entropy relationship provide putative relevant electrophysiological information to investigate the sensory-motor processing in normal condition and conditions such as movement disorders.

Neural code alterations and abnormal time patterns in Parkinson's disease

OBJECTIVE

The neural code used by the basal ganglia is a current question in neuroscience, relevant for the understanding of the pathophysiology of Parkinson's disease. While a rate code is known to participate in the communication between the basal ganglia and the motor thalamus/cortex, different lines of evidence have also favored the presence of complex time patterns in the discharge of the basal ganglia. To gain insight into the way the basal ganglia code information, we studied the activity of the globus pallidus pars interna (GPi), an output node of the circuit.

APPROACH

We implemented the 6-hydroxydopamine model of Parkinsonism in Sprague-Dawley rats, and recorded the spontaneous discharge of single GPi neurons, in head-restrained conditions at full alertness. Analyzing the temporal structure function, we looked for characteristic scales in the neuronal discharge of the GPi.

MAIN RESULTS

At a low-scale, we observed the presence of dynamic processes, which allow the transmission of time patterns. Conversely, at a middle-scale, stochastic processes force the use of a rate code. Regarding the time patterns transmitted, we measured the word length and found that it is increased in Parkinson's disease. Furthermore, it showed a positive correlation with the frequency of discharge, indicating that an exacerbation of this abnormal time pattern length can be expected, as the dopamine depletion progresses.

SIGNIFICANCE

We conclude that a rate code and a time pattern code can co-exist in the basal ganglia at different temporal scales. However, their normal balance is progressively altered and replaced by pathological time patterns in Parkinson's disease.

Multiple-time-scale framework for understanding the progression of Parkinson's disease

Parkinson's disease is marked by neurodegenerative processes that affect the pattern of discharge of basal ganglia neurons. The main features observed in the parkinsonian globus pallidus pars interna (GPi), a subdomain of the basal ganglia that is involved in the regulation of voluntary movement, are pathologically increased and synchronized neuronal activity. How these changes affect the implemented neuronal code is not well understood. Our experimental temporal structure-function analysis shows that in parkinsonian animals the rate-coding window of GPi neurons needed for the proper performance of voluntary actions is reduced. The model of the GPi network that we develop and discuss here reveals indeed that the size of the rate-coding window shrinks as the network activity increases and is expanded if the coupling strength among the neurons is increased. This leads to the novel interpretation that the pathological neuronal synchronization in Parkinson's disease in the GPi is the result of a collective attempt to counterbalance the shrinking of the rate-coding window due to increased activity in GPi neurons.

Neuronal Entropy Depends on the Level of Alertness in the Parkinsonian Globus Pallidus in vivo

A new working hypothesis of Parkinson's disease (PD) proposes to focus on the central role of entropy increase in the basal ganglia (BG) in movement disorders. The conditions necessary for entropy increase in vivo are, however, still not fully described. We recorded the activity of single globus pallidus pars interna neurons during the transition from deep anesthesia to full alertness in relaxed, head-restrained, control, and parkinsonian (6-hydroxydopamine-lesioned group-lesioned) rats. We found that during awakening from anesthesia, the variation of neuronal entropy was significantly higher in the parkinsonian than in the control group. This implies in our view that in PD the entropy of the output neurons of the BG varies dynamically with the input to the network, which is determined by the level of alertness. Therefore, entropy needs to be interpreted as a dynamic, emergent property that characterizes the global state of the BG neuronal network, rather than a static property of parkinsonian neurons themselves. Within the framework of the "entropy hypothesis," this implies the presence of a pathological feedback loop in the parkinsonian BG, where increasing the network input results in a further increase of neuronal entropy and a worsening of akinesia.

Non-linear dynamics in parkinsonism

Over the last 30 years, the functions (and dysfunctions) of the sensory-motor circuitry have been mostly conceptualized using linear modelizations which have resulted in two main models: the “rate hypothesis” and the “oscillatory hypothesis.” In these two models, the basal ganglia data stream is envisaged as a random temporal combination of independent simple patterns issued from its probability distribution of interval interspikes or its spectrum of frequencies respectively. More recently, non-linear analyses have been introduced in the modelization of motor circuitry activities, and they have provided evidences that complex temporal organizations exist in basal ganglia neuronal activities. Regarding movement disorders, these complex temporal organizations in the basal ganglia data stream differ between conditions (i.e., parkinsonism, dyskinesia, healthy control) and are responsive to treatments (i.e., l-DOPA, deep brain stimulation). A body of evidence has reported that basal ganglia neuronal entropy (a marker for complexity/irregularity in time series) is higher in hypokinetic state. In line with these findings, an entropy-based model has been recently formulated to introduce basal ganglia entropy as a marker for the alteration of motor processing and a factor of motor inhibition. Importantly, non-linear features have also been identified as a marker of condition and/or treatment effects in brain global signals (EEG), muscular activities (EMG), or kinetic of motor symptoms (tremor, gait) of patients with movement disorders. It is therefore warranted that the non-linear dynamics of motor circuitry will contribute to a better understanding of the neuronal dysfunctions underlying the spectrum of parkinsonian motor symptoms including tremor, rigidity, and hypokinesia.