Effect and Relationship of Gait on Subcortical Local Field Potentials in Parkinson's Disease: A Systematic Review

ObjectivesDevelopments in deep brain stimulation (DBS) technology have enabled the ability to detect local field potentials (LFPs) in Parkinson disease (PD). Gait dysfunction is one of the most prevalent deficits seen in PD. However, no consensus has been reached on the effect of gait on LFPs and the relationship between LFPs and clinical measures of gait. The objective of this systematic review was to synthesize existing research regarding the relationship between gait dysfunction and LFPs in PD.MethodsA systematic search of the literature yielded a total of ten articles, including 132 patients with PD, which met the criteria for inclusion.ResultsBeta frequency band measures showed low-to-strong correlation to clinical gait measures (r = ?0.50 to 0.82). Two studies found decreased beta power during gait; one found increased beta frequency peaks during gait; and one found higher beta power during dual-task gait than during single-task gait. One of the three studies comparing patients with and without freezing found significantly increased beta burst duration and power during gait in freezers compared with nonfreezers. All studies showed moderate-to-high methodologic quality.ConclusionsThis review highlights the need to consider the effect of gait on LFP recordings, particularly when used to guide DBS programming. Although sample sizes were small, it appears LFPs are associated to and modulated by gait in patients with PD. This evidence suggests that LFPs have the potential to be used as a biomarker of gait dysfunction in PD.     

Dopamine lesion-induced changes in subthalamic nucleus activity are not associated with alterations in firing rate or pattern in layer V neurons of the anterior cingulate cortex in anesthetized rats

Dysfunctional activity in the subthalamic nucleus (STN) is thought to underlie movement deficits of patients with Parkinson's disease. Alterations in STN firing patterns are also evident in the anesthetized rat model of Parkinson's disease, where studies show that loss of striatal dopamine and concomitant changes in the indirect pathway are associated with bursty and oscillatory firing patterns in STN output. However, the extent to which alterations in cortical activity contribute to changes in STN activity is unclear. As pyramidal neurons in the cingulate cortex project directly to the STN, cingulate output was assessed after dopamine lesion by simultaneously recording single-unit and local field potential (LFP) activities in STN and anterior cingulate cortex in control, dopamine-lesioned and non-lesioned hemispheres of urethane-anesthetized rats. Correlated oscillations were observed in cross-correlograms of spike trains from STN and cingulate layer V neurons with broad waveforms indicative of pyramidal neurons. One-2 weeks after dopamine cell lesion, firing rate, incidence of bursty and 0.3-2.5 Hz oscillatory activity of neurons and LFP power in the STN all increased significantly. In contrast, firing rate, incidence of bursty and 0.3-2.5 Hz oscillatory activity of cingulate layer V putative pyramidal neurons and power in cingulate LFPs did not differ significantly between dopamine-lesioned, non-lesioned or control hemispheres, despite significant loss of dopamine in the lesioned cingulate cortex. Data show that alterations in STN activity in the dopamine-lesioned hemisphere are not associated with alterations in neuronal activity in layer V of the anterior cingulate cortex in anesthetized rats.     

Motor unit synchronization in physiologic, enhanced physiologic, and voluntary tremor in man

Synchronization between pairs of single motor units simultaneously recorded from wrist extensor muscles was quantitated in 3 normal subjects during physiologic tremor (PT), beta-adrenergically enhanced physiologic tremor (EPT), and fast voluntary wrist flexion-extension movements mimicking tremor (VT). Cross-correlation histograms generated from the two spike trains of each motor unit pair demonstrated central or paracentral peaks in 13/19 recordings during PT, 22/36 during EPT, and 6/7 during VT. Relative peak area was used as a quantitative index of synchronization between the two motor units of each pair. It was lowest in PT, progressively increased in EPT as tremor amplitude increased, and highest in VT. In PT and lower amplitude EPT, the synchronization indexes were higher between motor units that discharged at the same or nearly the same frequency. In contrast, in higher amplitude EPT and VT, motor units with different firing frequencies were sometimes strongly synchronized as a consequence of double discharges in faster-firing motor units that had burst repetition rates in the range of slower-firing motor units discharging as singlets. Greater motor unit synchronization with increasing tremor amplitude in EPT may be secondary to a simultaneous increase in muscle spindle afferent activity from the tremulous muscle. Greatest synchronization in VT presumably reflects near maximal supraspinal and segmental common synaptic input onto motoneurons that generate VT. These results support a longstanding hypothesis that synchronization of motor units is the physiological basis for higher amplitude tremor.     

Single-trial estimation of neuronal firing rates: from single-neuron spike trains to population activity

We present a method to estimate the neuronal firing rate from single-trial spike trains. The method, based on convolution of the spike train with a fixed kernel function, is calibrated by means of simulated spike trains for a representative selection of realistic dynamic rate functions. We derive rules for the optimized use and performance of the kernel method, specifically with respect to an effective choice of the shape and width of the kernel functions. An application of our technique to the on-line, single-trial reconstruction of arm movement trajectories from multiple single-unit spike trains using dynamic population vectors illustrates a possible use of the proposed method.     

A nonparametric approach for detection of bursts in spike trains

In spike-train data, bursts are considered as a unit of neural information and are of potential interest in studies of responses to any sensory stimulus. Consequently, burst detection appears to be a critical problem for which the Poisson-surprise (PS) method has been widely used for 20 years. However, this method has faced some recurrent criticism about the underlying assumptions regarding the interspike interval (ISI) distributions. In this paper, we avoid such assumptions by using a nonparametric approach for burst detection based on the ranks of ISI in the entire spike train. Similar to the PS statistic, a “Rank surprise” (RS) statistic is extracted. A new algorithm performing an exhaustive search of bursts in the spike trains is also presented. Compared to the performances of the PS method on realizations of gamma renewal processes and spike trains recorded in cat auditory cortex, we show that the RS method is very robust for any type of ISI distribution and is based on an elementary formalization of the definition of a burst. It presents an alternative to the PS method for non-Poisson spike trains and is simple to implement.     

Synchronization of single motor units during voluntary contractions in the upper and lower extremities.

OBJECTIVE: To investigate motor unit synchronization in the time and frequency domains and compare the amount and nature of this synchronization between upper and lower extremity muscles in human subjects. METHODS: A total of 120 motor unit pairs from biceps brachii (BB), first dorsal interosseous (1DI), vastus medialis (VM), and tibialis anterior (TA) on the dominant side were analyzed and compared. Pairs of motor unit spike trains were recorded from two concentric needle electrodes inserted within these muscles in healthy volunteers. Subjects were instructed to maintain a weak isometric contraction of these muscles so that an individual motor unit recorded from each concentric needle discharged at a steady rate of approximately 10 impulses/s. Pairs of motor unit spike trains were cross-correlated in the time domain, and coherence analysis in the frequency domain was performed on the same spike train data. RESULTS: Synchronization was seen in all the muscles studied. Strength of motor unit synchronization, expressed as synchronization index (SI), was greater in 1DI muscles compared to other muscles (P<0.01). Coherence analysis revealed significant association between motor unit firings in the 1--5 and 25--30 Hz frequency ranges in all the muscles studied. The incidence of 25--30 Hz coherence peaks were found to be greater for 1DI muscles compared to other muscles. CONCLUSION: The above results suggest a possible role for corticospinal projections in producing pre-synaptic inputs responsible for synchronization of motor unit firings and 25--30 Hz coherence peaks.     

Detection of synchrony in the discharge of a population of neurons. II. Implementation and sensitivity of a synchronization index

This report describes the use of a synchronization index (Is; Hamm et al., 1985a) and its sensitivity to various forms and degrees of synchrony between spike trains. The dependence of the Is on signal-to-noise ratio, the number of synchronized spike trains and their degree of synchrony is shown in analog and digital simulations. These simulations and a comparison with peristimulus time histograms under conditions of induced synchrony reveal that the Is is a sensitive measure of synchronization in a population of spike trains.     

Time-resolved and time-scale adaptive measures of spike train synchrony

A wide variety of approaches to estimate the degree of synchrony between two or more spike trains have been proposed. One of the most recent methods is the ISI-distance which extracts information from the interspike intervals (ISIs) by evaluating the ratio of the instantaneous firing rates. In contrast to most previously proposed measures it is parameter free and time-scale independent. However, it is not well suited to track changes in synchrony that are based on spike coincidences. Here we propose the SPIKE-distance, a complementary measure which is sensitive to spike coincidences but still shares the fundamental advantages of the ISI-distance. In particular, it is easy to visualize in a time-resolved manner and can be extended to a method that is also applicable to larger sets of spike trains. We show the merit of the SPIKE-distance using both simulated and real data.     

Spike detection in human muscle sympathetic nerve activity using the kurtosis of stationary wavelet transform coefficients

The accurate assessment of autonomic sympathetic function is important in the diagnosis and study of various autonomic and cardiovascular disorders. Sympathetic function in humans can be assessed by recording the muscle sympathetic nerve activity, which is characterized by synchronous neuronal discharges separated by periods of neural silence dominated by colored Gaussian noise. The raw nerve activity is generally rectified, integrated, and quantified using the integrated burst rate or area. We propose an alternative quantification involving spike detection using a two-stage stationary wavelet transform (SWT) de-noising method. The SWT coefficients are first separated into noise-related and burst-related coefficients on the basis of their local kurtosis. The noise-related coefficients are then used to establish a threshold to identify spikes within the bursts. This method demonstrated better detection performance than an unsupervised amplitude discriminator and similar wavelet-based methods when confronted with simulated data of varying burst rate and signal to noise ratio. Additional validation on data acquired during a graded head-up tilt protocol revealed a strong correlation between the mean spike rate and the mean integrate burst rate (r = 0.85) and burst area rate (r = 0.91). In conclusion, the kurtosis-based wavelet de-noising technique is a potentially useful method of studying sympathetic nerve activity in humans.     

Spectral analysis of oscillatory neural circuits

Oscillatory dynamics are found at all levels of the nervous system. The goal of our current research on the control of rhythmic motor output by the lamprey spinal cord is to determine the features of neuronal coupling that lead to stable oscillatory activity and precisely-controlled intersegmental phase. Since our experimental manipulations can greatly increase the variability of the ventral root bursting pattern, it is important for us to employ a data analysis method which remains valid independent of this variability. Traditional analysis approaches which rely on identification of burst event times do not generally satisfy this requirement. In this paper, we illustrate the application of a straightforward statistically-based method for determining important parameters of oscillatory motor circuits using Fourier spectral analysis of spike trains. The frequency, phase, and their variabilities can be quantified; and the relative strength of coupling between different parts of the circuit can be tested for statistical significance. The approach we adopt is highly convenient for neuroscientists who study oscillatory systems as it operates directly on trains of action potentials stored as lists of event times (point-processes). Basic concepts and practical issues concerning use of Fourier analysis are discussed.     

Time-frequency analysis of transient neuromuscular events: dynamic changes in activity of the subthalamic nucleus and forearm muscles related to the intermittent resting tremor

In order to investigate the dynamic change in transient neuromuscular events and the functional correlation between the neural and muscular activity, local field potentials (LFPs) of the subthalamic nucleus (STN) and surface electromyograms (sEMGs) over several episodes of transient resting tremor from a patient with Parkinson's disease were quantitatively characterised in time-frequency domain using short-time Fourier transform and continuous wavelet transform. Events of onset and cessation of the tremor-related activity in the STN and muscles were correlated to reveal the temporal relationship between the two signals. A significant suppression in the power of the STN LFPs in the beta band (10-30 Hz) preceded the onset of resting tremor, which was presented as the increases in the power at the tremor frequency (3.0-4.5 Hz) in both STN LFPs and surface EMGs. Over the episodes of the intermittent resting tremor, the power of the STN LFPs in the beta band and the power of sEMGs in the tremor frequency band change in an alternating pattern with a significant exponential correlation (P(STN) = 16.8+62.3 x exp(-P(EMG)/6270.7); R2 = 0.72; p < 0.05). Significant linear correlation in the power values at the tremor frequency appears between STN LFPs and sEMGs (P(STN) = 65.1 + 2.1 x 10(-4)P(EMG); R2 = 0.41; p < 0.05). In comparison with short-time Fourier transform, similar results could be achieved using continuous wavelet transform of an appropriate wavelet with a higher temporal resolution but larger distortion in the high frequency.     

Wavelet-based processing of neuronal spike trains prior to discriminant analysis

Investigations of neural coding in many brain systems have focused on the role of spike rate and timing as two means of encoding information within a spike train. Recently, statistical pattern recognition methods, such as linear discriminant analysis (LDA), have emerged as a standard approach for examining neural codes. These methods work well when data sets are over-determined (i.e., there are more observations than predictor variables). But this is not always the case in many experimental data sets. One way to reduce the number of predictor variables is to preprocess data prior to classification. Here, a wavelet-based method is described for preprocessing spike trains. The method is based on the discriminant pursuit (DP) algorithm of Buckheit and Donoho [Proc. SPIE 2569 (1995) 540-51]. DP extracts a reduced set of features that are well localized in the time and frequency domains and that can be subsequently analyzed with statistical classifiers. DP is illustrated using neuronal spike trains recorded in the motor cortex of an awake, behaving rat [Laubach et al. Nature 405 (2000) 567-71]. In addition, simulated spike trains that differed only in the timing of spikes are used to show that DP outperforms another method for preprocessing spike trains, principal component analysis (PCA) [Richmond and Optican J. Neurophysiol. 57 (1987) 147-61].     

Detecting precise firing sequences in experimental data

A precise firing sequence (PFS) is defined here as a sequence of three spikes with fixed delays (up to some time accuracy Delta), that repeat excessively. This paper provides guidelines for detecting PFSs, verifying their significance through surrogate spike trains, and identifying existing PFSs. The method is based on constructing a three-fold correlation among spikes, estimating the expected shape of the correlation by smoothing, and detecting points for which the correlations significantly protrude above the expected correlation. Validation is achieved by generating surrogate spike trains in which the time of each of the real spikes is randomly jittered within a small time window. The method is extensively tested through application to simulated spike trains, and the results are illustrated with recordings of single units in the frontal cortex of behaving monkeys. Pitfalls which may cause false detection of PFSs, or loss of existing PFSs, include searching for PFSs in which the same neuron participates more than once, and attempting to produce a surrogate with some fixed statistical property.     

Oscillatory pallidal local field potential activity inversely correlates with limb dyskinesias in Parkinson's disease

Levodopa induced dyskinesias (LIDs) are poorly understood and yet are a major cause of disability in Parkinson's disease (PD). The activity of neurons in the basal ganglia of patients with PD tends to be strongly synchronized at frequencies under 30 Hz, leading to oscillatory local field potentials (LFPs). As dopaminergic therapy acutely suppresses this synchronization, we investigated whether this suppression may contribute to LIDs. Accordingly, we sought an inverse correlation between oscillatory synchronization and dyskinesia activity across time. To this end, we recorded pallidal LFPs in two Parkinsonian subjects exhibiting LIDs following surgery for deep brain stimulation. We correlated LFP power with simultaneously recorded EMG from the dyskinetic contralateral upper limb. We found highly significant inverse correlations between the oscillatory LFP activity under 30 Hz and dyskinetic EMG (maximum r = -0.65, P < 0.001 and r = -0.33, P < 0.001 for activities over 13-30 Hz in each subject). The inverse relationship between oscillatory pallidal LFP activity and dyskinetic EMG was maintained over time periods of a few seconds and was focal. This observation links the suppression of oscillatory synchronization in the pallidum with dyskinetic muscle activity in PD.     

A simple indicator of nonstationarity of firing rate in spike trains

The assessment of stationarity of firing rate in neural spike trains is important but is often performed only visually. Facing the growing amount of neural data generated by multi-electrode recording, there is a need for an automatic method to identify and disqualify spike trains with highly nonstationary firing rates. In this report, we propose a simple test of nonstationarity, associated with an indicator quantifying the degree of nonstationary in a spike train. This method is compared to the Mann-Kendall test of trend detection and the Runs test on simulated and real spike trains.     

How computational technique and spike train properties affect coherence detection

Spike train coherence is used to characterize common inputs that drive motor unit synchronization. However, data segmentation, overlap, and taper can affect coherence magnitude, thereby influencing the incidence at which significant coherence is detected. Also, the effect of spike train firing rate and common input variability on the detection of significant coherence is unknown. We used a pool of simulated synchronized spike trains with various firing rates (7-19 Hz), coefficients of variation (CV) (0.05-0.50), common input frequencies (10, 20, and 30 Hz, CV: 0.05-0.50), trial durations (30, 60, 90 and 120 s), and synchronization strength to explore the effects of segment length (1024 and 2048 1-ms samples), tapering (Hann, Nuttall, and rectangular), and overlap (0, 37.5, 50, 62.5, and 75%). Tapered segments overlapped by at least 50% maximized coherence, regardless of taper type. Coherence for 30-s trials revealed significant coherence for less than half of the motor unit pairs, demonstrating the advantages of longer trails. The 2048-sample segments produced similar coherence values with twice the frequency resolution. Increasing the common input variability from 0.15 to 0.50 reduced coherence incidence by approximately 60%, indicating that synchronized physiological motor unit pairs may fail to show significant coherence if the common input frequency is sufficiently unstable.     

Gamma oscillations in the auditory cortex of awake rats

Numerous reports of human electrophysiology have demonstrated gamma (30–150 Hz) frequency oscillations in the auditory cortex during listening. However, only a small number of studies in non‐human animals have provided evidence for gamma oscillations during listening. In this report, multi‐site recordings from primary auditory cortex (A1) were carried out using a 16‐channel microelectrode array in awake rats as they passively listened to tones. We addressed two fundamental questions: (i) Is passive listening associated with an increase in gamma oscillation in A1? And, if so: (ii) Are A1 gamma oscillations during passive listening coherent within local networks and/or over long distances? All sites within A1 showed a short‐latency burst of activity in the low‐gamma (30–70 Hz) and high‐gamma (90–150 Hz) bands in the local field potential (LFP). Additionally, 53% of sites within A1 also showed longer‐latency bursts of gamma oscillation that occurred episodically for up to 350 ms after tone onset, but these varied both in latency and in occurrence across trials. There was significant coherence in the low‐gamma band between spike activity and the LFP recorded with the same electrode. However, neither LFPs nor the spike activity between sites spaced at least 300 μm apart showed coherent activity in the gamma band. The experiments demonstrated that gamma oscillations are present, but not uniformly expressed, throughout A1 during passive listening and that there is strong local coherence in the spatiotemporal organization of gamma activity.