Noël L.W. Keijsers

Ambulatory motor assessment in Parkinson's disease

We developed an algorithm that distinguishes between on and off states in patients with Parkinsons 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 Parkinsons disease patients and can operate successfully in unsupervised ambulatory conditions.

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 Parkinsons 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.

Differential progression of proprioceptive and visual information processing deficits in Parkinson's disease.

Indirect evidence suggests that patients with Parkinsons disease (PD) have deficits not only in motor performance, but also in the processing of sensory information. We investigated the role of sensory information processing in PD patients with a broad range of disease severities and in a group of age‐matched controls. Subjects were tested in two conditions: pointing to a remembered visual target in complete darkness (DARK) and in the presence of an illuminated frame with a light attached to the index finger (FRAME). Differences in pointing errors in these two conditions reflect the effect of visual feedback on pointing. PD patients showed significantly larger constant and variable errors than controls in the DARK and FRAME condition. The difference of the variable error in the FRAME and DARK condition decreased as a function of the severity of PD. This indicates that any deficits in the processing of proprioceptive information occur already at very mild symptoms of PD, and that deficits in the use of visual feedback develop progressively in later stages of the disease. These results provide a tool for early diagnosis of PD and shed new light on the functional role of the brain structures that are affected in PD.

Online monitoring of dyskinesia in patients with Parkinson's disease

Discusses utilizing wearable movement sensors to successfully detect and assess severity of Parkinsonion symptoms in daily life. The authors describe the present state-of-the-art of online monitoring of motor behavior. They focus on modern monitoring equipment and on new analysis tools, which allow a quantitative detection and assessment of dyskinesias throughout the day, by an objective, unsupervised rating of dyskinesias using clinical rating scores.

A new method to normalize plantar pressure measurements for foot size and foot progression angle

Plantar pressure measurement provides important information about the structure and function of the foot and is a helpful tool to evaluate patients with foot complaints. In general, average and maximum plantar pressure of 6-11 areas under the foot are used to compare groups of subjects. However, masking the foot means a loss of important information about the plantar pressure distribution pattern. Therefore, the purpose of this study was to develop and test a simple method that normalizes the plantar pressure pattern for foot size, foot progression angle, and total plantar pressure. Moreover, scaling the plantar pressure to a standard foot opens the door for more sophisticated analysis techniques such as pattern recognition and machine learning. Twelve subjects walked at preferred and half of the preferred walking speed over a pressure plate. To test the method, subjects walked in a straight line and in an approaching angle of approximately 40 degrees . To calculate the normalized foot, the plantar pressure pattern was rotated over the foot progression angle and normalized for foot size. After normalization, the mean shortest distance between the contour lines of straight walking and walking at an angle had a mean of 0.22 cm (SD: 0.06 cm) for the forefoot and 0.14 cm (SD: 0.06 cm) for the heel. In addition, the contour lines of normalized feet for the various subjects were almost identical. The proposed method appeared to be successful in aligning plantar pressure of various feet without losing information.

Detection and assessment of the severity of Levodopa-induced dyskinesia in patients with Parkinson's disease by neural networks

Levodopa‐induced dyskinesias (LID) in Parkinsons disease (PD) have remained a clinical challenge. We evaluated the feasibility of neural networks to detect LID and to quantify their severity in 16 patients with PD at rest and during various activities of daily living. The movements of the patients were measured using four pairs of accelerometers mounted on the wrist, upper arm, trunk, and leg on the most affected side. Using parameters obtained from the accelerometer signals, neural networks were trained to detect and to classify LID corresponding to the modified Abnormal Involuntary Movement Scale. Important parameters for classification appeared to be the mean segment velocity and the cross‐correlation between accelerometers on the arm, trunk, and leg. Neural networks were able to distinguish voluntary movements from LID and to assess the severity of LID in various activities. Based on the results in this study, we conclude that neural networks are a valid and reliable method to detect and to assess the severity of LID corresponding to the modified Abnormal Involuntary Movement Scale.

Plantar pressure with and without custom insoles in patients with common foot complaints.

Background: Although many patients with foot complaints receive customized insoles, the choice for an insole design can vary largely among foot experts. To investigate the variety of insole designs used in daily practice, the insole design and its effect on plantar pressure distribution were investigated in a large group of patients. Materials and Methods: Mean, peak, and pressure-time-integral per sensor for 204 subjects with common foot complaints for walking with and without insoles was measured with the footscan® insole system (RSscan International). Each insole was scanned twice (precision3D), after which the insole height along the longitudinal and transversal cross section was calculated. Subjects were assigned to subgroups based on complaint and medial arch height. Data were analyzed for the total group and for the separate subgroups (forefoot or heel pain group and flat, normal or high medial arch group). Results: The mean pressure significantly decreased under the metatarsal heads II-V and the calcaneus and significantly increased under the metatarsal bones and the lateral foot (p < 0.0045) due to the insoles. However, similar redistribution patterns were found for the different foot complaints and arch heights. There was a slight difference in insole design between the subgroups; the heel cup was significantly higher and the midfoot support lower for the heel pain group compared to the forefoot pain group. The midfoot support was lowest in the flat arch group compared to the high and normal arch group (p < 0.05). Conclusion: Although the insole shape was specific for the kind of foot complaint and arch height, the differences in shape were very small and the plantar pressure redistribution was similar for all groups. Clinical Relevance: This study indicates that it might be sufficient to create basic insoles for particular patient groups.

Classification of forefoot pain based on plantar pressure measurements

BACKGROUND Plantar pressure is widely used to evaluate foot complaints. However, most plantar pressure studies focus on the symptomatic foot with foot deformities. The purposes of this study were to investigate subjects without clear foot deformities and to identify differences in plantar pressure pattern between subjects with and without forefoot pain. The second aim was to discriminate between subjects with and without forefoot pain based on plantar pressure measurements using neural networks. METHODS In total, 297 subjects without foot deformities of whom almost 50% had forefoot pain walked barefoot over a pressure plate. Foot complaints and subject characteristics were assessed with a questionnaire and a clinical evaluation. Plantar pressure was analyzed using a recently developed method, which produced pressure images of the time integral, peak pressure, mean pressure, time of activation and deactivation, and total contact time per pixel. After pre-processing the pressure images with principal component analysis, a forward selection procedure with neural networks was used to classify forefoot pain. FINDINGS The pressure-time integral and mean pressure were significantly larger under the metatarsals II and III for subjects with forefoot pain. A neural network with 14 input parameters correctly classified forefoot pain in 70.4% of the test feet. INTERPRETATION The differences in plantar pressure parameters between subjects with and without forefoot pain were small. The reasonable performance of forefoot pain classification by neural networks suggests that forefoot pain is related more to the distribution of the pressure under the foot than to the absolute values of the pressure at fixed locations.

Flat Feet, Happy Feet? Comparison of the Dynamic Plantar Pressure Distribution and Static Medial Foot Geometry between Malawian and Dutch Adults

In contrast to western countries, foot complaints are rare in Africa. This is remarkable, as many African adults walk many hours each day, often barefoot or with worn-out shoes. The reason why Africans can withstand such loading without developing foot complaints might be related to the way the foot is loaded. Therefore, static foot geometry and dynamic plantar pressure distribution of 77 adults from Malawi were compared to 77 adults from the Netherlands. None of the subjects had a history of foot complaints. The plantar pressure pattern as well as the Arch Index (AI) and the trajectory of the center of pressure during the stance phase were calculated and compared between both groups. Standardized pictures were taken from the feet to assess the height of the Medial Longitudinal Arch (MLA). We found that Malawian adults: (1) loaded the midfoot for a longer and the forefoot for a shorter period during roll off, (2) had significantly lower plantar pressures under the heel and a part of the forefoot, and (3) had a larger AI and a lower MLA compared to the Dutch. These findings demonstrate that differences in static foot geometry, foot loading, and roll off technique exist between the two groups. The advantage of the foot loading pattern as shown by the Malawian group is that the plantar pressure is distributed more equally over the foot. This might prevent foot complaints.

Reliability of the revised wheelchair rugby trunk impairment classification system

Study design:Observational, cross-sectional.Objectives:A new classification system for trunk impairment in wheelchair rugby was introduced in 2010. It consists of 10 tests, arranged in an algorithm, to assign four different trunk scores (0, 0.5, 1.0 or 1.5) to athletes. The purpose of this study was to assess the inter-rater reliability of this classification system.Setting:National competition for wheelchair rugby and wheelchair basketball in the Netherlands and Belgium.Methods:Three experienced wheelchair rugby classifiers independently assigned trunk scores to wheelchair rugby and wheelchair basketball athletes in two sessions. After each session, test descriptions were adjusted. The inter-rater reliability was evaluated by determining the agreement and Fleiss Kappa.Results:In the first session, all classifiers agreed on the trunk score in 13 out of 16 athletes; the overall Kappa was 0.76 (P<0.001). The Kappa per trunk score ranged from 0.29 to 1. Four test descriptions were adjusted after the first session. In the second session, there was an agreement in trunk score between the classifiers in 15 out of 21 athletes. The overall Kappa was 0.75 (P<0.0001), and the Kappa per trunk scores ranged from 0.58 to 0.92. After the second session, two test descriptions were improved.Conclusion:The revised classification system for trunk impairment in wheelchair rugby showed a adequate inter-rater reliability for the allocation of trunk scores.