Accelerometric assessment of levodopa-induced dyskinesias in Parkinson's disease

Automated Levodopa-induced dyskinesia assessment

An automated methodology for Levodopa-induced dyskinesia (LID) assessment is presented in this paper. The methodology is based on the analysis of the signals recorded from accelerometers and gyroscopes, which are placed on certain positions on the subject's body. The obtained signals are analyzed and several features are extracted. Based on these features a classification technique is used for LID detection and classification of its severity. The method has been evaluated using a group of 10 subjects. Results are presented related to each individual sensor as well as for various sensor combinations. The obtained results indicate high classification ability (93.73% classification accuracy).

Monitoring motor fluctuations in patients with Parkinson's disease using wearable sensors

This paper presents the results of a pilot study to assess the feasibility of using accelerometer data to estimate the severity of symptoms and motor complications in patients with Parkinson's disease. A support vector machine (SVM) classifier was implemented to estimate the severity of tremor, bradykinesia and dyskinesia from accelerometer data features. SVM-based estimates were compared with clinical scores derived via visual inspection of video recordings taken while patients performed a series of standardized motor tasks. The analysis of the video recordings was performed by clinicians trained in the use of scales for the assessment of the severity of Parkinsonian symptoms and motor complications. Results derived from the accelerometer time series were analyzed to assess the effect on the estimation of clinical scores of the duration of the window utilized to derive segments (to eventually compute data features) from the accelerometer data, the use of different SVM kernels and misclassification cost values, and the use of data features derived from different motor tasks. Results were also analyzed to assess which combinations of data features carried enough information to reliably assess the severity of symptoms and motor complications. Combinations of data features were compared taking into consideration the computational cost associated with estimating each data feature on the nodes of a body sensor network and the effect of using such data features on the reliability of SVM-based estimates of the severity of Parkinsonian symptoms and motor complications.

Using wearable sensors to predict the severity of symptoms and motor complications in late stage Parkinson's Disease

This paper is focused on the analysis of data obtained from wearable sensors in patients with Parkinson's Disease. We implemented Support Vector Machines (SVM's) to predict clinical scores of the severity of Parkinsonian symptoms and motor complications. We determined the optimal window length to extract features from the sensor data. Furthermore, we performed tests to determine optimal parameters for the SVM's. Finally, we analyzed how well individual tasks performed by patients captured the severity of various symptoms and motor complications.

Joint angle tracking with inertial sensors

Many wearable inertial systems have been used to continuously track human movement in and outside of a laboratory. The number of sensors and the complexity of the algorithms used to measure position and orientation vary according to the clinical application. To calculate changes in orientation, researchers often integrate the angular velocity. However, a relatively small error in measured angular velocity leads to large integration errors. This restricts the time of accurate measurement to a few minutes. We have combined kinematic models designed for control of robotic arms with state space methods to directly and continuously estimate the joint angles from inertial sensors. These algorithms can be applied to any combination of sensors, can easily handle malfunctions or the loss of some sensor inputs, and can be used in either a real-time or an off-line processing mode with higher accuracy.

Analyzing 180 degrees turns using an inertial system reveals early signs of progression of Parkinson's disease

Changes in turning are one of the early motor deficiencies in Parkinson's Disease (PD). We have proposed a system based on wearable, inertial sensors and a novel automatic analysis algorithm that can assess 180 degrees turns. Twelve patients in early stages of PD and 14 age-matched healthy subjects were enrolled in this study. Inertial sensors were attached on shanks and sternum. Measurement protocol included walking on a straight pathway, turning 180 degrees and returning back. Subjects were measured 4 times, once every 6 months during an 18 months period. At the baseline, 9 subjects from each group repeated the test twice to assess test-retest reliability. Patients with mild PD had a very low Postural Instability Gait Difficulty (PIGD subscore of UPDRS III) score (average 0.67, min 0, max 3). The analysis showed that the patients had a significantly longer turning duration (2.18+/-0.43 vs. 1.79+/-0.27 seconds, p<0.02) and longer delay in their last step before initiating a turn (0.56+/-0.04 vs. 0.52+/-0.04 seconds, p<0.03). Estimated turning duration and other metrics had a high test-retest reliability (rho>0.85). Turning duration also showed a significant Group *Time interaction (p<0.03) during the longitudinal study highlighting early signs of the progression of the disease.

Long-term monitoring of gait in Parkinson's disease

A new system for long-term monitoring of gait in Parkinson's disease (PD) has been developed and validated. The characteristics of every stride taken over 10-h epochs were acquired using a lightweight ankle-mounted sensor array that transmitted data wirelessly to a small pocket PC at a rate of 100 Hz. Stride was calculated from the vertical linear acceleration and pitch angular velocity of the leg with an accuracy of 5 cm. Results from PD patients (5) demonstrate the effectiveness of long-term monitoring of gait in a natural environment. The small, variable stride length characteristic of Parkinsonian gait, and fluctuations of efficacy associated with levodopa therapy, such as delayed onset, wearing off, and the 'off/on' effect, could reliably be detected from long-term changes in stride length.

The relations among putative biorisk markers in schizotypal adolescents: minor physical anomalies, movement abnormalities, and salivary cortisol

BACKGROUND

Evidence suggests that prenatal insult may play a role in the etiology of psychotic disorders. Minor physical anomalies (MPA) are an indicator of abnormal fetal development and are elevated in individuals at genetic and behavioral risk for psychosis. Yet, there has been little empirical research on the relationships between MPAs and other neurobiological risk indicators. We hypothesized that the frequency of MPAs (an external marker of prenatal central nervous system [CNS] disruption) would be associated with two other biomarkers suggestive of disruptions in fetal neurodevelopment: movement abnormalities (an indicator of striatal abnormalities) and heightened cortisol secretion (an indicator of hypothalamic-pituitary-adrenal [HPA]/hippocampal function).

METHODS

Participants with schizotypal personality disorder (SPD; n = 39) and both normal (n = 47) and other personality disorders (n = 28) control subjects were administered structured diagnostic interviews and assessed for MPAs, movement abnormalities, and salivary cortisol.

RESULTS

Schizotypal personality disorder participants showed significantly greater MPAs and movement abnormalities and higher cortisol than both the normal and other personality disorders groups. Hierarchical linear regression analyses revealed that higher rates of MPAs were linked with greater movement abnormalities and salivary cortisol.

CONCLUSIONS

The findings suggest that MPAs serve as a marker of neurodevelopmental abnormalities that affect striatal and hippocampal regions.

Quantification of Tremor and Bradykinesia in Parkinson's Disease Using a Novel Ambulatory Monitoring System

An ambulatory system for quantification of tremor and bradykinesia in patients with Parkinson's disease (PD) is presented. To record movements of the upper extremities, a sensing units which included miniature gyroscopes, has been fixed to each of the forearms. An algorithm to detect and quantify tremor and another algorithm to quantify bradykinesia have been proposed and validated. Two clinical studies have been performed. In the first study, 10 PD patients and 10 control subjects participated in a 45-min protocol of 17 typical daily activities. The algorithm for tremor detection showed an overall sensitivity of 99.5% and a specificity of 94.2% in comparison to a video reference. The estimated tremor amplitude showed a high correlation to the Unified Parkinson's Disease Rating Scale (UPDRS) tremor subscore (e.g., r = 0.87, p < 0.001 for the roll axis). There was a high and significant correlation between the estimated bradykinesia related parameters estimated for the whole period of measurement and respective UPDRS subscore (e.g., r = -0.83, p < 0.001 for the roll axis). In the second study, movements of upper extremities of 11 PD patients were recorded for periods of 3-5 hr. The patients were moving freely during the measurements. The effects of selection of window size used to calculate tremor and bradykinesia related parameters on the correlation between UPDRS and these parameters were studied. By selecting a window similar to the period of the first study, similar correlations were obtained. Moreover, one of the bradykinesia related parameters showed significant correlation (r = -0.74, p < 0.01) to UPDRS with window sizes as short as 5 min. Our study provides evidence that objective, accurate and simultaneous assessment of tremor and bradykinesia can be achieved in free moving PD patients during their daily activities.

Monitoring Activity in Individuals with Parkinson Disease

This study was conducted to investigate the validity of an Activity Monitor (AM) recording functional activities in individuals with Parkinson disease (PD). A series of tasks were performed in both a fixed and a random manner by 11 participants with PD (H & Y 2-3; age range 40-79 years). Participants, wearing an AM, were recorded for video analysis (VA). The strength of association (rs) between the AM and VA for duration of time in body positions, walking, and bicycling across both fixed and random modules ranged from 0.63 to 0.98, however, the AM reported significantly greater time spent in body positions in both modules. Kappa statistics (K) between the AM and VA were the highest for transitions in the fixed module with Ks ranging from 0.74 to 1.0. During the random module, the agreement between the AM and VA was lowest for transitions between sitting and standing. In general, the Ks were low for activities lasting less than 5 seconds. The data collected in the study support the conclusion that the AM accurately evaluated the duration of time spent in body positions, the number of transitions between body positions, and the duration of time spent walking for activities lasting longer than 5 seconds in individuals with PD.

Ambulatory motor assessment in Parkinson's disease

We developed an algorithm that distinguishes between on and off states in patients with Parkinson's disease during daily life activities. Twenty-three patients were monitored continuously in a home-like situation for approximately 3 hours while they carried out normal daily-life activities. Behavior and comments of patients during the experiment were used to determine the on and off periods by a trained observer. Behavior of the patients was measured using triaxial accelerometers, which were placed at six different positions on the body. Parameters related to hypokinesia (percentage movement), bradykinesia (mean velocity), and tremor (percentage peak frequencies above 4 Hz) were used to distinguish between on and off states. The on-off detection was evaluated using sensitivity and specificity. The performance for each patient was defined as the average of the sensitivity and specificity. The best performance to classify on and off states was obtained by analysis of movements in the frequency domain with a sensitivity of 0.97 and a specificity of 0.97. We conclude that our algorithm can distinguish between on and off states with a sensitivity and specificity near 0.97. This method, together with our previously published method to detect levodopa-induced dyskinesia, can automatically assess the motor state of Parkinson's disease patients and can operate successfully in unsupervised ambulatory conditions.

Quantifying drug-induced dyskinesias in the arms using digitised spiral-drawing tasks

In this study, we quantify the severity of drug-induced dyskinesias in the arms of Parkinson's disease (PD) patients using digitised spiral-drawing tasks. Two spiral drawings, namely a circular and a square spiral, are designed to, respectively, represent the continuous and discrete arm motions, and the size of the spiral is decided so that both the distal and proximal arm joints are involved. Fifteen PD patients, average disease duration 14.4+/-7.4 years, were assessed 30 min after a levodopa challenge whilst performing circular and square spiral-drawing tasks. The velocity of drawing movements was computed and the amplitude of the involuntary dyskinetic movements was measured as the standard deviation of the drawing velocity (SD-DV). The mean amplitude of dyskinetic movements was compared between arms and tasks and was correlated with clinical measures including the Bain dyskinesia scale and the total unified Parkinson's disease rating scale (UPDRS) score. The results showed that there was no statistically significant difference in the amplitude of dyskinesias either between the arms or between the continuous (circular) and discrete (square) spiral drawings in this group of PD patients, but interestingly the interaction between arm and drawing pattern was significant. Significant correlations were found between the magnitude of dyskinesia measured from the spiral-drawing tasks and both the 'on' or 'off' UPDRS and also the Bain dyskinesia scale. We conclude that the drawing tasks may be used to provide an objective method of quantifying the severity of drug-induced dyskinesias in the arm in PD patients.

The use of the Actiwatch–Neurologica® system to objectively assess the involuntary movements and sleep–wake activity in patients with mild–moderate Huntington’s disease

Huntington's disease (HD) is a neurodegenerative disorder characterised by cognitive, psychiatric and motor abnormalities including a range of involuntary movements. Currently, assessment of these movements involves the use of subjective rating scales such as the Unified Huntington's Disease Rating Scales (UHDRS) for bradykinesia and maximal dystonia and chorea, without any objective measures. As new therapies emerge, it is critical that an objective means of evaluating these abnormal movements is developed and we have investigated the use of a wrist-worn activity monitor, the Actiwatch-Neurologica, to determine whether these movements can be measured. In addition, this activity monitor and subjective reports were used to objectively measure the degree of sleep disruption in these same HD patients. Eight patients with mild-moderate HD and 8 age- and sex-matched control subjects wore the monitor for a period of 48 hours and recorded in a diary whether they were asleep or awake for each hour over the 2-day period. Assessment of various movement parameters revealed that HD patients exhibited significantly greater total and maximum activity levels and spent longer performing high acceleration movements while they were awake compared with controls. During sleep, patients not only showed significantly more activity and spent more time making high acceleration movements, but they also made significantly more movements than control subjects. These results demonstrate that the Actiwatch-Neurologica activity monitor can be used to objectively assess movements in HD patients during periods of high activity as well as during sleep.

Accelerometric measurement of involuntary movements during pallidal deep brain stimulation of patients with generalized dystonia

Accelerometric activity during rest and posture was quantified in the upper dominant limb of 14 patients with primary or secondary dystonia and five healthy control subjects. Data were recorded before and after bilateral implantation of the stimulating electrodes in the Globus Pallidus internus. Clinical evaluation was based on the Burke-Marsden-Fahn's Dystonia Rating Scale (BMFDRS). For the patient group, I(t), the integral (i.e. area) of the acceleration power spectrum over the total frequency range (0.6-16 Hz) decreased as the clinical state of the patients improved following deep brain stimulation (p < 0.01) during rest and posture. Ten days after surgery, there were no I(t) differences between control subjects and patients (p > 0.05). A significant correlation was found between the global BMFDRS scores and I(t) for rest (p < 0.01) but not for posture. No significant correlation was found between I(t) and a partial BMFDRS score for the right arm for rest or posture. The integral I(t) provides a valid indicator of the motor activity generated by the arm of the patient but further analyses are needed to monitor patients' progress not only during their hospitalization but also after they are released from the hospital, and to understand why this measure does not correlate with partial BMFDRS scores.

Accuracy of Objective Ambulatory Accelerometry in Detecting Motor Complications in Patients With Parkinson Disease

Shortcomings of existing assessment methods in Parkinson disease (PD) have led to the development of continuous ambulatory multichannel accelerometry for the assessment of the core features of PD. Although measures for hypokinesia, bradykinesia, and tremor have been validated in groups of patients with PD, it is unclear whether this method is able to detect "on" with or without dyskinesias, and "off" in individual PD patients. This study therefore addressed the accuracy of objective ambulatory accelerometry in detecting motor complications in 15 PD patients, using a self-assessment scale as gold standard. Measures for hypokinesia, bradykinesia, and tremor showed limited sensitivity (0.60-0.71) and specificity (0.66-0.76) for motor complications in individual PD patients. In the group of PD patients, comparing the "on" with the "off" state yielded statistically significant differences for tremor only. Objective dyskinesia measures correlated with time spent with dyskinesias (r = 0.89). Although validated for the measurement of hypokinesia, bradykinesia, and tremor, continuous ambulatory multichannel accelerometry currently cannot detect "on" and "off" in individual PD patients.

Movement parameters that distinguish between voluntary movements and levodopa-induced dyskinesia in Parkinson's disease

It is well known that long-term use of levodopa by patients with Parkinson's disease causes dyskinesia. Several methods have been proposed for the automatic, unsupervised detection and classification of levodopa induced dyskinesia. Recently, we have demonstrated that neural networks are highly successful to detect dyskinesia and to distinguish dyskinesia from voluntary movements. The aim of this study was to use the trained neural networks to extract parameters, which are important to distinguish between dyskinesia and voluntary movements. Thirteen patients were continuously monitored in a home-like situation performing in about 35 daily life tasks for a period of approximately 2.5 h. Behavior of the patients was measured using triaxial accelerometers, which were placed at six different positions of the body. A neural network was trained to assess the severity of dyskinesia. The neural network was able to assess the severity of dyskinesia and could distinguish dyskinesia from voluntary movements in daily life. For the trunk and the leg, the important parameters appeared to be the percentage of time that the trunk or leg was moving and the standard deviation of the segment velocity of the less dyskinetic leg. For the arm, the combination of the percentage of time, that the wrist was moving, and the percentage of time, that a patient was sitting, explained the largest part of the variance of the output. Dyskinesia differs from voluntary movements in the fact that dyskinetic movements tend to have lower frequencies than voluntary movements and in the fact that movements of different body segments are not well coordinated in dyskinesia.

Automatic assessment of levodopa-induced dyskinesias in daily life by neural networks

We developed an objective and automatic procedure to assess the severity of levodopa-induced dyskinesia (LID) in patients with Parkinson's disease during daily life activities. Thirteen patients were continuously monitored in a home-like situation for a period of approximately 2.5 hours. During this time period, the patients performed approximately 35 functional daily life activities. Behavior of the patients was measured using triaxial accelerometers, which were placed at six different positions on the body. A neural network was trained to assess the severity of LID using various variables of the accelerometer signals. Neural network scores were compared with the assessment by physicians, who evaluated the continuously videotaped behavior of the patients off-line. The neural network correctly classified dyskinesia or the absence of dyskinesia in 15-minute intervals in 93.7, 99.7, and 97.0% for the arm, trunk, and leg, respectively. In the few cases of misclassification, the rating by the neural network was in the class next to that indicated by the physicians using the AIMS score (scale 0-4). Analysis of the neural networks revealed several new variables, which are relevant for assessing the severity of LID. The results indicate that the neural network can accurately assess the severity of LID and could distinguish LID from voluntary movements in daily life situations.

Force overflow and levodopa-induced dyskinesias in Parkinson's disease

We assessed force coordination of the hand in Parkinson's disease and its relationship to motor complications of levodopa therapy, particularly to levodopa-induced dyskinesias (LID). We studied two groups of Parkinson's disease patients with (Parkinson's disease + LID, n = 23) and without levodopa-induced dyskinesias (Parkinson's disease - LID, n = 10), and age-matched healthy controls. The motor score of the Unified Parkinson's Disease Rating Scale, a dyskinesia score and force in a grip-lift paradigm were assessed ON and OFF levodopa. A pathological increase of forces was seen in ON-state in Parkinson's disease + LID only. In Parkinson's disease + LID, the force involved in pressing down the object before lifting was significantly increased by levodopa (by 61%, P < 0.05). An overshooting of peak grip force by 51% (P < 0.05) and of static grip force by 45% (P < 0.01) was observed in the ON- compared with the OFF-drug condition. In contrast, no excessive force was found in Parkinson's disease - LID. Peak grip force in ON-state was 140% (P < 0.05) higher in Parkinson's disease + LID than in Parkinson's disease - LID, while static grip force was increased by 138% (P < 0.01) between groups. Severity of peak-dose dyskinesias was strongly correlated with grip force in ON-state (r = 0.79 with peak force, P < 0.01). No correlation was observed between forces and the motor score as well as with the daily dose of dopaminergic medication. Force excess was only observed in patients with LID and motor fluctuations. A close relationship was seen between the overshooting of forces and dyskinesias in the ON-drug condition. We postulate that both LID and grip force excess share common pathophysiological mechanisms related to motor fluctuations.