Comparing kinematic changes between a finger-tapping task and unconstrained finger flexion-extension task in patients with Parkinson's disease

Deep learning of Parkinson's movement from video, without human-defined measures

BACKGROUND

The core clinical sign of Parkinson's disease (PD) is bradykinesia, for which a standard test is finger tapping: the clinician observes a person repetitively tap finger and thumb together. That requires an expert eye, a scarce resource, and even experts show variability and inaccuracy. Existing applications of technology to finger tapping reduce the tapping signal to one-dimensional measures, with researcher-defined features derived from those measures.

OBJECTIVES

(1) To apply a deep learning neural network directly to video of finger tapping, without human-defined measures/features, and determine classification accuracy for idiopathic PD versus controls. (2) To visualise the features learned by the model.

METHODS

152 smartphone videos of 10s finger tapping were collected from 40 people with PD and 37 controls. We down-sampled pixel dimensions and videos were split into 1 s clips. A 3D convolutional neural network was trained on these clips.

RESULTS

For discriminating PD from controls, our model showed training accuracy 0.91, and test accuracy 0.69, with test precision 0.73, test recall 0.76 and test AUROC 0.76. We also report class activation maps for the five most predictive features. These show the spatial and temporal sections of video upon which the network focuses attention to make a prediction, including an apparent dropping thumb movement distinct for the PD group.

CONCLUSIONS

A deep learning neural network can be applied directly to standard video of finger tapping, to distinguish PD from controls, without a requirement to extract a one-dimensional signal from the video, or pre-define tapping features.

Parkinsonian Hand or Clinician's Eye? Finger Tap Bradykinesia Interrater Reliability for 21 Movement Disorder Experts

BACKGROUND

Bradykinesia is considered the fundamental motor feature of Parkinson's disease (PD). It is central to diagnosis, monitoring, and research outcomes. However, as a clinical sign determined purely by visual judgement, the reliability of humans to detect and measure bradykinesia remains unclear.

OBJECTIVE

To establish interrater reliability for expert neurologists assessing bradykinesia during the finger tapping test, without cues from additional examination or history.

METHODS

21 movement disorder neurologists rated finger tapping bradykinesia, by Unified Parkinson's Disease Rating Scale (MDS-UPDRS) and Modified Bradykinesia Rating Scale (MBRS), in 133 videos of hands: 73 from 39 people with idiopathic PD, 60 from 30 healthy controls. Each neurologist rated 30 randomly-selected videos. 19 neurologists were also asked to judge whether the hand was PD or control. We calculated intraclass correlation coefficients (ICC) for absolute agreement and consistency of MDS-UPDRS ratings, using standard linear and cumulative linked mixed models.

RESULTS

There was only moderate agreement for finger tapping MDS-UPDRS between neurologists, ICC 0.53 (standard linear model) and 0.65 (cumulative linked mixed model). Among control videos, 53% were rated >  0 by MDS-UPDRS, and 24% were rated as bradykinesia by MBRS subscore combination. Neurologists correctly identified PD/control status in 70% of videos, without strictly following bradykinesia presence/absence.

CONCLUSION

Even experts show considerable disagreement about the level of bradykinesia on finger tapping, and frequently see bradykinesia in the hands of those without neurological disease. Bradykinesia is to some extent a phenomenon in the eye of the clinician rather than simply the hand of the person with PD.

Unprompted Alteration of Freely Chosen Movement Rate During Stereotyped Rhythmic Movement: Examples and Review

Investigations of behavior and control of voluntary stereotyped rhythmic movement contribute to the enhancement of motor function and performance of disabled, sick, injured, healthy, and exercising humans. The present article presents examples of unprompted alteration of freely chosen movement rate during voluntary stereotyped rhythmic movements. The examples, in the form of both increases and decreases of movement rate, are taken from activities of cycling, finger tapping, and locomotion. It is described that, for example, strength training, changed power output, repeated bouts, and changed locomotion speed can elicit an unprompted alteration of freely chosen movement rate. The discussion of the examples is based on a tripartite interplay between descending drive, rhythm-generating spinal neural networks, and sensory feedback, as well as terminology from dynamic systems theory.

Synergic control of action in levodopa-naïve Parkinson's disease patients: II. Multi-muscle synergies stabilizing vertical posture

Postural instability is a major disabling feature in Parkinson's disease (PD). We quantified the organization of leg and trunk muscles into synergies stabilizing the center of pressure (COP) coordinate within the uncontrolled manifold hypothesis in levodopa-naïve patients with PD and age-matched control subjects. The main hypothesis was that changes in the synergic control of posture are present early in the PD process even before levodopa exposure. Eleven levodopa-naïve patients with PD and 11 healthy controls performed whole-body cyclical voluntary sway tasks and a self-initiated load-release task during standing on a force plate. Surface electromyographic activity in 13 muscles on the right side of the body was analyzed to identify muscle groups with parallel scaling of activation levels (M-modes). Data were collected both before ("off-drug") and approximately 60 min after the first dose of 25/100 carbidopa/levodopa ("on-drug"). COP-stabilizing synergies were quantified for the load-release task. Levodopa-naïve patients with PD showed no COP-stabilizing synergy "off-drug", whereas controls showed posture-stabilizing multi-M-mode synergy. "On-drug", patients with PD demonstrated a significant increase in the synergy index. There were no significant drug effects on the M-mode composition, anticipatory postural adjustments, indices of motor equivalence, or indices of COP variability. The results suggest that levodopa-naïve patients with PD already show impaired posture-stabilizing multi-muscle synergies that may be used as promising behavioral biomarkers for emerging postural disorders in PD. Moreover, levodopa modified synergy metrics differently in these levodopa-naïve patients compared to a previous study of patients on chronic antiparkinsonian medications (Falaki et al. in J Electromyogr Kinesiol 33:20-26, 2017a), suggesting different neurocircuitry involvement.

Contralateral Transfer of the Phenomenon of Repeated Bout Rate Enhancement in Unilateral Index Finger Tapping

These hypotheses were tested: (1) Freely chosen frequency in unilateral index finger tapping is correlated between the two index fingers, and (2) A 3-min bout of unilateral index finger tapping followed by 10 min rest results in an increase of the freely chosen tapping frequency performed by the contralateral index finger in a second bout. Thirty-two adults participated. Freely chosen tapping frequencies from first bouts were 167.2 ± 79.0 and 161.5 ± 69.4 taps/min for the dominant and non-dominant hand, respectively (p=.434). These variables correlated (R=.86, p<.001). When bout one and two were performed with the dominant and non-dominant hand, respectively, the frequency increased by 8.1%±17.2% in bout two (p=.011). In opposite order, the frequency increased by 14.1%±17.5% (p<.001), which was not different from the ∼8% (p=.157).

Force Increase in a Repetitive Motor Task Inducing Motor Fatigue

To evaluate task induced motor fatigue in a well-established finger tapping task, we analyzed tapping parameters and included the time course of measures of force. We hypothesized that a decline in tapping force would reflect task induced motor fatigue, defined by a lengthening of inter-tap intervals (ITI). A secondary aim was to investigate the reliability of tapping data acquisition with the force sensor. Results show that, as expected, tapping speed decreased linearly over time, due to both an increase of ITI and tap duration. In contrast, tapping force increased non-linearly over time and was uncorrelated to changes in tapping speed. Force data could serve as a measure to characterize task induced motor fatigue. Force sensors can assess a decline in tapping speed as well as an independent increase of tapping force. We argue that the increase of force reflects central compensation, i.e. perception of fatigue, due to an increase in task effort and difficulty.

Repeated Bout Rate Enhancement Is Elicited by Various Forms of Finger Tapping

Voluntary rhythmic movements, such as, for example, locomotion and other cyclic tasks, are fundamental during everyday life. Patients with impaired neural or motor function often take part in rehabilitation programs, which include rhythmic movements. Therefore, it is imperative to have the best possible understanding of control and behaviour of human voluntary rhythmic movements. A behavioural phenomenon termed repeated bout rate enhancement has been established as an increase of the freely chosen index finger tapping frequency during the second of two consecutive tapping bouts. The present study investigated whether the phenomenon would be elicited when the first bout consisted of imposed passive finger tapping or air tapping. These two forms of tapping were applied since they can be performed without descending drive (passive tapping) and without afferent feedback related to impact (air tapping) – as compared to tapping on a surface. Healthy individuals (n = 33) performed 3-min tapping bouts separated by 10 min rest. Surface electromyographic, kinetic, and kinematic data were recorded. Supportive experiments were made to measure, for example, the cortical sensory evoked potential (SEP) response during the three different forms of tapping. Results showed that tapping frequencies in the second of two consecutive bouts increased by 12.9 ± 14.8% (p < 0.001), 9.9 ± 6.0% (p = 0.001), and 16.8 ± 13.6% (p = 0.005) when the first bout had consisted of tapping, passive tapping, and air tapping, respectively. Rate enhancement occurred without increase in muscle activation. Besides, the rate enhancements occurred despite that tapping, as compared with passive tapping and air tapping, resulted in different cortical SEP responses. Based on the present findings, it can be suggested that sensory feedback in an initial bout increases the excitability of the spinal central pattern generators involved in finger tapping. This can eventually explain the phenomenon of repeated bout rate enhancement seen after a consecutive bout of finger tapping.

A 12-Week Cycling Training Regimen Improves Gait and Executive Functions Concomitantly in People with Parkinson's Disease

Background: There is increasing evidence that executive functions and attention are associated with gait and balance, and that this link is especially prominent in older individuals or those who are afflicted by neurodegenerative diseases that affect cognition and/or motor functions. People with Parkinson’s disease (PD) often present gait disturbances, which can be reduced when PD patients engage in different types of physical exercise (PE), such as walking on a treadmill. Similarly, PE has also been found to improve executive functions in this population. Yet, no exercise intervention investigated simultaneously gait and non-motor symptoms (executive functions, motor learning) in PD patients.

Objective: To assess the impact of aerobic exercise training (AET) using a stationary bicycle on a set of gait parameters (walking speed, cadence, step length, step width, single and double support time, as well as variability of step length, step width and double support time) and executive functions (cognitive inhibition and flexibility) in sedentary PD patients and healthy controls.

Methods: Two groups, 19 PD patients (Hoehn and Yahr ≤2) and 20 healthy adults, matched on age and sedentary level, followed a 3-month stationary bicycle AET regimen.

Results: Aerobic capacity, as well as performance of motor learning and on cognitive inhibition, increased significantly in both groups after the training regimen, but only PD patients improved their walking speed and cadence (all p < 0.05; with no change in the step length). Moreover, in PD patients, training-related improvements in aerobic capacity correlated positively with improvements in walking speed (r = 0.461, p < 0.05).

Conclusion: AET using stationary bicycle can independently improve gait and cognitive inhibition in sedentary PD patients. Given that increases in walking speed were obtained through increases in cadence, with no change in step length, our findings suggest that gait improvements are specific to the type of motor activity practiced during exercise (i.e., pedaling). In contrast, the improvements seen in cognitive inhibition were, most likely, not specific to the type of training and they could be due to indirect action mechanisms (i.e., improvement of cardiovascular capacity). These results are also relevant for the development of targeted AET interventions to improve functional autonomy in PD patients.

Examining associations between psychosis risk, social anhedonia, and performance of striatum-related behavioral tasks

Both psychosis and anhedonia have been associated to some extent with striatal functioning. The current study examined whether either psychosis risk or social anhedonia was associated with performance on 3 tasks related to striatal functioning. Psychosis risk participants had extremely elevated Perceptual Aberration/Magical Ideation (PerMag) scores (n = 69), with 43% of psychosis risk participants also having semistructured interview-assessed psychotic-like experiences which further heightens their risk of psychotic disorder (Chapman, Chapman, Kwapil, Eckblad, & Zinser, 1994). Compared with both extremely elevated social anhedonia (n = 60) and control (n = 68) groups, the PerMag group exhibited poorer performance on 2 of the striatum-related tasks, the Weather Prediction Task (WPT) and the Learned Irrelevance Paradigm, but not on Finger Tapping. In addition, PerMag participants with psychotic-like experiences were especially impaired on the WPT. Overall, this study arguably provides the first evidence that psychosis risk but not social anhedonia is associated with performance on the WPT, a task thought to be strongly associated with activation in the associative striatum, and also suggests that the WPT might be especially useful as a behavioral measure of psychosis risk.

Contribution of Step Length to Increase Walking and Turning Speed as a Marker of Parkinson's Disease Progression

When increasing ambulation speed in Parkinson’s disease, step cadence increases more than stride length, indicating movement scaling difficulties that affect step generation in particular. We investigated whether step length variation when increasing ambulation speed was related to disease progression. Patients with Parkinson’s disease (N = 39) and controls (N = 152) performed two timed ambulation tasks: at a 'free' (self-selected) pace and then at 'maximal' speed. The total number of steps (including during turns) and time to complete the task were clinically measured. The relative contribution of step length and cadence to increased ambulation speed was determined using two methods: the ratios of change in step length or in cadence to the change in ambulation speed, and the step length index. While the relative contribution of step length and cadence to increased ambulation speed was independent of age in both control and patient groups, in Parkinson’s disease there was a negative correlation between time from diagnosis and the ratio of change in step length to change in ambulation speed (R = 0.54; p = 0.0004) and the step length index (R = 0.56, p = 0.0002). In parallel, there was a positive correlation between time since diagnosis and the ratio of change in cadence to change in ambulation speed (R = 0.57; p = 0.0002). The relative contribution of step length and cadence to increased ambulation speed is age invariant but a marker of Parkinson's disease advancement, and can be easily determined in the clinical setting.

Prehension synergies and hand function in early-stage Parkinson's disease

We explored the multi-digit synergies and hand performance in object manipulations and pressing tasks in patients with early-stage Parkinson's disease (PD) and healthy controls. Synergies were defined as inter-trials co-variation patterns among forces/moments produced by individual digits that stabilized a resultant mechanical variable. The subjects performed three main tasks: pressing (steady-state force production followed by a force pulse into the target), prehension (manipulation of a handheld instrumented handle imitating the action of taking a sip from a glass), and functional object manipulation (moving a glass with water as quickly and accurately as possible along a chain of targets). The PD patients were slower compared to controls in all three tasks. Patients showed smaller synergy indices in the pressing and prehension tasks. In the prehension tasks, patients showed elevated grip force at steady states with smaller grip force modulation during the handle motion. PD patients showed smaller feed-forward synergy adjustments in preparation to the quick action in the pressing and (to a smaller degree) prehension tasks. Synergy indices correlated with the time index of performance in the functional glass-with-water task, whereas none of the indices correlated with the Unified PD Rating Scale part III-motor scores. We interpret the results as pointing at an important role of subcortical structures in motor synergies and their feed-forward adjustments to action.

Impact of regional striatal dopaminergic function on kinematic parameters of Parkinson's disease

Among the cardinal parkinsonian motor deficits, the severity of bradykinesia correlates with striatal dopamine loss. However, the impact of regional striatal dopamine loss on specific components of bradykinesia remains unknown. Using gyroscopes, we measured the amplitude, speed, and frequency of finger tapping in 24 untreated patients with Parkinson's disease (PD) and 28 healthy controls. Using positron emission tomography (PET) studies and [(18)F]-N-3-fluoropropyl-2-beta-carboxymethoxy-3-beta-(4-iodophenyl) nortropane (FP-CIT) in PD patients, we investigated the relationship between the mean values, variability and decrements of various kinematic parameters of finger tapping on one side (e.g. the mean, variability and decrement) and contralateral striatal FP-CIT binding. Compared with controls, PD patients had reduced amplitudes and speeds of tapping and showed greater decrement in those parameters. PD patients also exhibited greater irregularity in amplitude, speed, and frequency. Putaminal FP-CIT uptake levels correlated with the mean speed and amplitude, and caudate uptake levels correlated with mean amplitude. The variability of amplitude and speed correlated only with the caudate uptake levels. Neither caudate nor putaminal uptake correlated with frequency-related parameters or decrement in amplitude or speed. Reduced amplitude and speed of repetitive movement may be related to striatal dopaminergic deficit. Dopaminergic action in the caudate nucleus is required to maintain consistency of amplitude and speed. Although decrement of amplitude and speed is known to be specific for PD, we found that it did not mirror the degree of striatal dopamine depletion.