Skip to main content

Characterization of stroke-related upper limb motor impairments across various upper limb activities by use of kinematic core set measures

Abstract

Background

Upper limb kinematic assessments provide quantifiable information on qualitative movement behavior and limitations after stroke. A comprehensive characterization of spatiotemporal kinematics of stroke subjects during upper limb daily living activities is lacking. Herein, kinematic expressions were investigated with respect to different movement types and impairment levels for the entire task as well as for motion subphases.

Method

Chronic stroke subjects with upper limb movement impairments and healthy subjects performed a set of daily living activities including gesture and grasp movements. Kinematic measures of trunk displacement, shoulder flexion/extension, shoulder abduction/adduction, elbow flexion/extension, forearm pronation/supination, wrist flexion/extension, movement time, hand peak velocity, number of velocity peaks (NVP), and spectral arc length (SPARC) were extracted for the whole movement as well as the subphases of reaching distally and proximally. The effects of the factors gesture versus grasp movements, and the impairment level on the kinematics of the whole task were tested. Similarities considering the metrics expressions and relations were investigated for the subphases of reaching proximally and distally between tasks and subgroups.

Results

Data of 26 stroke and 5 healthy subjects were included. Gesture and grasp movements were differently expressed across subjects. Gestures were performed with larger shoulder motions besides higher peak velocity. Grasp movements were expressed by larger trunk, forearm, and wrist motions. Trunk displacement, movement time, and NVP increased and shoulder flexion/extension decreased significantly with increased impairment level. Across tasks, phases of reaching distally were comparable in terms of trunk displacement, shoulder motions and peak velocity, while reaching proximally showed comparable expressions in trunk motions. Consistent metric relations during reaching distally were found between shoulder flexion/extension, elbow flexion/extension, peak velocity, and between movement time, NVP, and SPARC. Reaching proximally revealed reproducible correlations between forearm pronation/supination and wrist flexion/extension, movement time and NVP.

Conclusion

Spatiotemporal differences between gestures versus grasp movements and between different impairment levels were confirmed. The consistencies of metric expressions during movement subphases across tasks can be useful for linking kinematic assessment standards and daily living measures in future research and performing task and study comparisons.

Trial registration: ClinicalTrials.gov Identifier NCT03135093. Registered 26 April 2017, https://clinicaltrials.gov/ct2/show/NCT03135093.

Introduction

The human upper limb can be expressed by seven main degrees of freedom, excluding the hand, that allow for highly variable movements and interactions with the environment. After stroke, this movement complexity can be affected due to a disruption in the cerebral sensorimotor networks that lead to inefficient or abnormal movement activation [1]. Sensitive assessments of the motor function on the level of movement quality and influences of deficits on daily life functionality are important to reveal relevant movement limitations and drive interventions for improving functional restoration [2].

Over the last decades, upper limb kinematic assessments have increasingly been used as primary or secondary outcome measures, next to standard clinical assessments in randomized-controlled trials concerning stroke rehabilitation [3,4,5,6,7,8,9,10,11,12,13,14,15,16]. Upper limb kinematic assessments have been investigated to test the effectiveness of different therapies, namely constraint induced movement therapy [14,15,16], trunk restraint training [17], robotic-assisted training [18,19,20], virtual reality training [21, 22], bilateral arm training [23,24,25], Botulinum toxin [26] and mirror therapy [27]. This tendency demonstrates the additional value of kinematic assessments to complement the standard clinical assessments and their broad evaluation level of movement quality. The advantage of upper limb kinematic measurements, compared to standard clinical assessments, is that different aspects of motion can be tracked objectively and continuously [28]. Alongside with this emerging field, the variability and heterogeneity of kinematic assessment protocols and chosen outcomes increased, making it difficult to interpret findings across studies [29]. A systematic review of upper limb kinematic assessments and metrics in subjects after stroke published criteria regarding upper limb assessment protocols. Out of 151 different metrics, task / movement time, path length ratio, number of velocity peaks, shoulder flexion/extension angle, trunk displacement, and peak velocity were proposed as core set metrics for facilitating the standardization and comparability of upper limb kinematic analysis after stroke [30]. These core set metrics performed best in terms of usage frequency, validity and/ or reliability. It has further been shown that upper limb kinematic measurements after stroke were frequently assessed in relatively fixed measurement surroundings such as camera-based motion laboratories or robot-based measurement systems. These measurements have the strong disadvantage that the movements the patients have to perform are device-specific and often restricted to simple reach-to-point, or tracking motions [30]. It is questionable, to what extent, a device-restraint planar pointing task is representative for movement tasks in daily living. It is unknown, if the movement characteristics of these different tasks are different for the same person as well as between people performing the same task or not. A significant impact of the movement task content and contexts on motor planning and behavior has been reported for pointing, or grasping of simulated or real objects [31, 32] and might be one of the biggest barriers in the interpretation and overall comparison of upper limb movement kinematics. With the exception of some natural tasks such as the standardized task of drinking from a glass [33], it is unknown how the outcomes of the current kinematic assessments relate to real-life performance and other similar functional upper limb movement tasks.

To overcome the described issue of highly variable and complex movements of the human upper limb, effort has been put into the development of a taxonomy for upper limb motion that subdivides motions based on the complexity and duration into activities, functional movements, and functional primitives [34]. The functional movement primitives or movement subphases, such as reaching or transporting, were suggested to be seen as building blocks or even more granular elements of motion, that are consistent across movements [35]. Based on this theorem, it could be assumed that for example reaching to grasp a glass is based on the same primitive or building block as reaching to grasp a phone receiver. Observing upper limb motions on the level of the functional movement primitives or movement subphases could thereby enable across task comparisons of movement quality and overcome issues of anatomical and task-related complexity. In person with stroke-related upper limb impairments, the kinematic analysis on the level of movement subphases could thus help to uncover pathophysiological mal-adaptations and relevant limitations in movement behavior such as diminished elbow extension during reach or arm elevation and hand speed during object transport.

Most of the pre-mentioned research on upper limb kinematic assessments was based on one or two movement tasks narrowing the findings down to a specific type of movement. A wider set of upper limb assessment activities, including non-contact movements such as gestures, or contact movements such as grasping activities, would increase the representativeness and comprehensiveness of the kinematic characterization of upper limb movement quality in daily life. In the present study, a set of 20 movement tasks representative for activities of daily life was used covering the main requirements of movement control of the human upper limb degrees of freedom (DOF) in terms of workspace, grasp configuration, interaction with the environment and complexity [36]. The aim of the study is to characterize and differentiate stroke-related upper limb function and impairment for movements related to activities of daily life. To this end, kinematics presenting the main upper limb spatiotemporal movement characteristics were recorded during gesture and grasping actions.

The first question is whether kinematic characteristics are different between movement types of no-contact-based gesture movements and contact-based grasping. The second question is whether significant effects can be found related to subgroups of no, mild and moderate upper limb impairment. The third question is attributed to the comparability of movement subphases. It is questioned whether phases of reaching distally towards ipsilateral maximum arm length and reaching or transporting proximally towards the head are consistent in terms of spatiotemporal kinematic expressions and relations across different movement tasks and impairment levels.

Methods

A prospective cross-sectional observational study on subjects after chronic stroke and healthy subjects was performed to explore the relationship between upper limb function and activities as measured by clinical assessments and by a wearable sensor-based motion capture system. The study took place between July 2017 and October 2019 at the rehabilitation clinic cereneo (Vitznau, Switzerland).

Study participants

The study sample consisted of 26 subjects with a unilateral ischemic or hemorrhagic stroke in the chronic stage (> 6 months) with presence of partial upper limb motor impairment, nonetheless with the ability to lift the arm against gravity (> 30° of shoulder flexion) and to flex and extend the fingers for basic grasp performances. The subjects were excluded, if increased upper limb muscle tone inducing limitations in range of motion (modified Ashworth Scale ≥ 3 in one of the tested muscle groups), severe sensory deficits in the upper limb (Erasmus modifications to the revised Nottingham Sensory Assessment of 0 in one of the test regions), or upper limb impairments unrelated to the neurological disease, such as preexisting orthopedic problems were present. Five age-matched healthy subjects without history of neurologic impairments nor limitations in upper limb movements were included for the acquisition of kinematic reference data. Each participant had to be able to understand and follow basic commands to perform the study experiments and to give written informed consent before inclusion, according to the Declaration of Helsinki and the Swiss regulatory authorities (BASEC-ID: 2016-02075).

Study experiments

All experiments were performed by an experienced research therapist during a single-day measurement per subject at the rehabilitation clinic cereneo (Vitznau, Switzerland). After donning and calibration of the inertial sensor system, each participant was asked to perform a set of upper limb activities three times. Stroke participants were asked to perform the movements with the affected upper limb, healthy participants used their non-dominant side.

During the experiment study participants were sitting in upright position on an armless chair. The start and end position for the gesture movements were defined with the tested arm drooped straight with the hand approximately above the hip. The grasping movements were performed on a height-adjustable table with the start and end position of the tested arm defined by 90° of elbow flexion and neutral shoulder position. The corresponding position of the index fingers on the table was marked as a reference for each subject. After finishing the task, the participant was asked to return to the start position.

The experimental task selection is based on previous works [36, 37], consisting of ten intransitive, gesture movements, and ten transitive, reach-to-grasp, and manipulation movements, as illustrated in Fig. 1 and described in detail in Table 1. This task set covers the main upper limb movement workspace and contains the main grasping types [38], while enabling the differentiation between gesture movements without contact and grasping actions with object contact and manipulation.

Fig. 1
figure 1

Experimental protocol of 20 activities of daily life. The task items T1–T10 include intransitive gesture movements without object contact. The task items T11–T20 represent transitive grasping movements with objects contact and manipulation. Tasks that were included in the subphase analysis of reaching distally are encircled in blue and those that include reaching proximally are encircled in green

Table 1 Set of daily living tasks

From these 20 tasks, a subset of tasks with similar workspace demands was chosen to test similarities of motion primitives across tasks. We defined and tested four primitives, (1) reaching distally to grasp, (2) reaching distally to gesture, (3) reaching proximally to transport, and (4) reaching proximally to gesture. As described by Schambra et al. reaches are defined by motions with the intention to make contact with an object or target that might include contact in terms of grasp or touch at the end of movement [34]. The motion primitive of transport is defined by the purpose to convey an object that can result in motion away from the body or towards the body or head specifically [34]. In the present study, relevant workspace directions were further differentiated, such as reaching towards maximum armlength reaching distance and reaching towards the head, as indicated in the task representations in Fig. 1. Subphases of reaching towards an object placed in ipsilateral arm length distance and transporting towards the mouth, ear, or head can be detected in the tasks, drinking from a glass (T12, T14), taking the phone for a call (T13), biting into an apple (T15), putting on a hat (T16), and tooth brushing (T18). Therefore, recorded kinematics of reaching distally to grasp (1) and proximally to transport (3) were investigated across the tasks, T12, T13, T14, T15, T16, and T18. The associated tasks to reaching to gesture distally (2) are the thumb down (T02), greeting (T06), and the stop gesture (T08). Reaching proximally to gesture (4) is tested in the tasks, protecting the face from the sun (T05), the military salute (T07), and the silence gesture (T10).

Measurement system

For primary outcome measures, kinematic data was recorded by use of a full-body, wearable motion capture system, Xsens MVN Awinda (Xsens Technologies, Enschede, The Netherlands). The system offers real-time visualization, playback and editing of human motion capture data by a set of 17 wireless sensors that were attached symmetrically onto predefined body-parts of the participant. The sensors included in the upper limb motion analysis of this study were limited to the upper body, located above the sternum, the shoulder blade, the upper arm, the forearm and the back of the hand, as illustrated in Fig. 2.

Fig. 2
figure 2

Upper limb sensor set-up. The location of the wearable sensors on predefined body segments is shown in relation to the global frame

Each sensor contains 3D accelerometry, 3D gyroscopes and 3D magnetometers to calculate subject-specific hand workspace and joint angles and positions [39]. The motion axes are pre-defined with respect to the global frame, where X is defined along the sagittal plane, Y along the horizontal plane and Z along the fontal plane in or against the direction of gravity as shown in Fig. 2. Positive values indicate outward, forward, or upward directions and negative values inward, backward, or downward directions. The upper limb kinematics captured with the Xsens MVN Awinda, included the maximum reach distance and movement range in vertical direction that have shown to correlate with the Fugl-Meyer Assessment of Upper Extremity [40]. The kinematic parameters were also applied in everyday surroundings for providing objective measurements of upper limb movements [41, 42].

Kinematic core set

The kinematic core set is comprised of 10 parameters, namely trunk displacement, shoulder flexion/extension, shoulder abduction/adduction, elbow flexion/extension, forearm pronation/ supination, wrist flexion/extension, movement time, peak velocity, the number of velocity peaks (NVP), and spectral arc length (SPARC). These 10 parameters are evaluated for the complete movement tasks and the four motion primitives (1–4) of reach and transport in the subject-specific workspace separately. Matlab (MATLAB version 2018b, The Mathwork, Natick, MA, USA) was used for data processing.

The range of motion in the shoulder, elbow and hand is defined by the scalar measure from minimum to maximum joint angle. As suggested in the MVN Xsens manual, the ZXY Euler rotation sequence is used to present the joint angles, where flexion/extension is defined around the Z-axis, abduction/adduction is defined around the X-axis and internal/external rotation is defined around the Y-axis. However, an exception is made for shoulder abduction/adduction angle, which is recommended to be read out by the Euler rotation sequence XZY, to reduce estimation errors due to gimbal lock [43, 44]. Shoulder flexion/extension is defined as the angle along the sagittal plane, determined by rotation of the upper arm around the z-axis of the Euler sequence ZXY along the sagittal plane. Shoulder abduction/adduction is defined by rotation of the upper arm around the y-axis of the Euler sequence XZY along the frontal plane. Elbow flexion/extension is defined by rotation of the lower arm around the x-axis. Forearm pronation/supination is defined around the y-axis and wrist flexion/extension by around the around the x-axis. Trunk compensation is measured by changes in position and orientation of the sternum sensor [45]. The changes were calculated by the Euclidean distance minus an offset given by the mean of the first 10 data points of the x-, y- and z-direction.

Outcomes focusing on temporal aspects of motion included movement time, speed, and smoothness metrics. Movement time was defined as the time between movement onset and movement end, by applying a threshold of 2% of the peak velocity [33] measured with the hand-sensor. Peak velocity is determined by the maximum of linear hand velocity in m/s along the three directions with respect to the global reference frame, that were summed up by square root of the sum of the three directions. Similarly, NVP were summed up for three directions, reflecting changes between acceleration and deceleration phases and thereby the smoothness of the movement profile. A velocity peak is defined as the data point that is larger than its two neighboring samples in the linear hand velocity profile. The NVP is applied dimensionless without a per time unit, and valid, however could lack sensitivity and reliability in case of measurement noise [46]. SPARC has been suggested for its robustness against measurement noise. SPARC was defined to reflect the spectral energy induced by unsmooth, saccadic motions [46].

Classification of relevant movement phases or motion primitives

As described in previous studies, segmentation of upper limb activities is important, to explore relevant aspects of task performance [33, 47, 48]. For the purpose of this study, the movement primitives, reach distally to grasp or gesture and reaching proximally to transport or gesture were further analyzed. Based on pre-assumption of a similar workspace across tasks, the phase segmentation was performed for the gesture movements T02, T05, T06, T07, T08, T10, and the grasp movements of the tasks T12, T13, T14, T15, T16, and T18 as highlighted in Fig. 1. Semi-automated phase detection was used by feature-based movement detection algorithms as exemplified in Fig. 3 for one trial of the drinking task (T12) of one subject. Similarly, feature-based segmentation has been used based on finger force detection and finger angular motions in reach-to-grasp movements [48], as well as based sensor signals of orientation angles in ambulation assessments [49].

Fig. 3
figure 3

Feature-based movement phase segmentation for one subject and a single trial. One trial of the drinking task is presented by dimensionless plotting of the hand IMU position and velocity signals. The red lines indicate the threshold detection of velocity in x-direction. The blue dashed line refers to threshold detection of velocity in z-direction

The movement subphases of reaching distally to grasp or gesture towards maximum ipsilateral arm length were determined based on the maximum distance in position in x-direction along the sagittal plane and the velocity data of the hand-IMU. The subphases of reaching proximally to transport or gesture in direction towards the head were determined by the maximum position of the hand-IMU in z-direction along gravity vector, as well as the minimum position of the hand-IMU in x-direction. Reaching or transporting distally towards maximum ipsilateral armlength is detected from movement onset to maximum distance of the hand-IMU in x-direction, combined with a velocity threshold in x- and z-direction. Reaching or transporting proximally towards the head was defined by the hand-IMU trajectory maximum height in z-direction combined with a velocity threshold in the x- and z-direction. The automatic detection method focused on detecting the highest/lowest peaks in the velocity profile of the hand-IMU and identifying the beginning of increase in absolute velocity according to a threshold of 2% of the peak velocity in m/s [33]. The start of the ‘Reach’ corresponds to the first increase of the velocity of the hand IMU in the x-direction. The following point corresponds to the increase in the velocity of the sensor in the z direction, indicating the beginning of the movement towards the head (‘Transport to head’). The start of the ‘Return Object’ phase is identified by the increase of the velocity in the negative z-direction. Finally, the last phase (‘Return to start’) is identified via the last negative peak of the velocity profile of the sensor’s x-direction. For gesture movement, only three points were determined, the movement onset, target position, as defined by maximum hand distance x-direction and or maximum height in z-direction, and the movement end. The automatic phase-detection was verified by visual inspection one trial after another and corrected manually if needed.

Statistical analysis

The statistical analysis was performed using Matlab (MATLAB version 2018b, The Mathwork, Natick, MA, USA) and SPSS (SPSS version 26.0, IBM Corp.,Armonk, N.Y., USA). All kinematic outcome parameters were tested for normal distribution by inspecting the histograms. Descriptive statistics of the kinematic outcome parameters were presented in mean and the 95% confidence intervals.

The first and second research question were addressed by performing a linear mixed effects model analysis for each kinematic outcome parameter to test for the impact of the factor gesture versus grasping movements and the factor of the impairment level as fixed effects. For the analysis of impairment-based expressions in kinematic outcomes, the participants were allocated to the no, mild or moderate impairment group based on the Fugl-Meyer Assessment of the Upper Extremity (FMMA-UE). The healthy subjects with a FMMA-UE full score of 66/66 points constituted the no impairment group. Subjects with a score ranging from 48 to 65 points were assigned to the mild impairment group and those with a score ranging from 32 to 47 points to the moderate impairment group [34, 50,51,52,53].

To address the third research question on comparability of kinematic expressions during movement subphases across different tasks and impairment subgroups, we considered the metrics during the subphases of reaching distally and proximally that were extracted from the tasks T02, T05, T06, T07, T08, T10, T12, T13, T14, T15, T16, and T18. Differences in metric expressions between tasks were tested with the Kruskal–Wallis test, a non-parametric version of the one-factor ANOVA for the subphases, reaching distally to grasp (1), reaching distally to gesture (2), reaching proximally to transport (3) and reaching proximally to gesture (4). Non-significant results in the Kruskal–Wallis test (p > 0.05) indicate that metric characteristics across tasks are similar. Additionally, correlations between the ten core set metrics of reaching distally and reaching proximally were explored visually in a correlation matrix for each task and group of impairment severity. Correlation coefficients of r ≥ 0.5 or r ≤ − 0.5 were defined as significant and compared across tasks and impairment severity.

Results

Thirty-one subjects were included in the present analysis, with the participant characteristics presented in Table 2. Fourteen subjects were affected on their dominant hand. Three subjects missed one movement task item (S09 missed T03, S13 missed T09, and S23 missed T17) of the 20 prescribed actions. The whole dataset consisted of 775 kinematic sets of the affected upper limb, represented by the mean values of the three repetitions per subject and task that has been published online [54].

Table 2 Participant characteristics

Kinematic core set of the total task execution

As shown in Table 3, all kinematic parameters were statistically significantly different between gesture and grasp movements when considering the total task execution across subjects. Gestures resulted in larger shoulder joint motions and higher peak velocity when compared to the grasping movements. Grasping movements were associated with larger trunk motions, increased elbow and wrist flexion/extension ranges and forearm pronation/supination as well as an increased NVP.

Table 3 Linear mixed model results on kinematics for the factor movement task across subjects

On the level of subgroup comparisons between subjects with no, mild, and moderate impairments, each kinematic metric was included in a linear mixed model with the results indicated in Fig. 4. Significant effects of the impairment level were found for trunk displacement (p = 0.010) and shoulder flexion/extension (p = 0.001) with statistically significant post-hoc comparison between no impairment and moderate impairment, as well as between mild and moderate impairment for both kinematic outcomes. The interactions between the task and the impairment significantly influenced measures of shoulder abduction/adduction (p = 0.037). Movement time was significantly affected by the factor’s impairment level (p < 0.001), when the affected side is the dominant side (p = 0.038), and the interaction between the task and the impairment level (p < 0.001). Subjects of the no impairment group performed tasks faster with a mean of 4.2 (3.3–5.1) seconds, the mild impairment group with a mean of 5.2 (4.6–5.7) seconds and the moderate impairment group with a mean of 6.4 (5.9–7.0) seconds. In post-hoc analysis, significant differences were computed between the no and moderate impairment group (p < 0.001) as well as between the mild and moderate impairment group (p = 0.006). The NVP were shown to be influenced by the factor impairment level (p = 0.001) and the interaction between the task and the impairment group (p = 0.002). Significantly larger NVP were found in the moderate impairment group with a mean of 53.0 (45.5–60.4) when compared to the no impairment group (p = 0.012) with a mean of 31.1 (18.8–43.3) and between the moderate impairment group and the mild impairment group (p = 0.003) with a mean of 34.5 (27.0–41.9).

Fig. 4
figure 4

Effects of the task and impairment group on core set kinematics. Abd/Add abduction/adduction, Flex/Ext flexion/extension, Pro/Sup pronation/supination, Mov movement, SPARC spectral arc length, TrunkDisp trunk displacement, Vel velocity. *Indicates significant effects between the no, mild, and/or moderate impairment group for both gesture and grasp movements

Core set kinematics for subphases of reaching proximally and distally with and without contact

Four classes of sub phases, reaching distally to grasp (1), reaching distally to gesture (2), reaching proximally to transport (3), and reaching proximally to gesture (4), were predefined and detected in the tasks T02, T05, T06, T07, T08, T10, T12, T13, T14, T15, T16, and T18. The automatic movement phase segmentation was verified without manual correction in 85% of the trials. Figure 5A illustrates the kinematic metric expressions while reaching distally towards the maximum reaching arm length (1, 2). The data is illustrated in one boxplot per task across subjects. Figure 5B represents the metrics during reaching proximally towards the person’s head (3, 4).

Fig. 5
figure 5figure 5

A Kinematic metrics during reaching distally per task across subjects (N = 31). The metric expressions are presented in boxplots per task. The vertical black line in each subplot separates the grasp movement tasks on the left side and gesture movement tasks on the right side. Statistically significant comparable metrics across tasks are indicated by asterisks and the bold notation of the gesture or grasping task caption. El-Flex/Ext elbow flexion/extension, FA-Pro/Sup forearm pronation/supination, Mov Time movement time, No of Vel Peak number of peak velocity, Sh-Abd/Add shoulder abduction/adduction, Sh-Flex/Ext shoulder flexion/extension, SPARC spectral arc length, TrunkDisp trunk displacement, Wr-Flex/Ext wrist flexion/extension. B Kinematic metrics during reaching proximally per task across subjects (N = 31). The metric expressions are presented in boxplots per task. The vertical black line in each subplot separates the grasp movement tasks on the left side and gesture movement tasks on the right side. Statistically significant comparable metrics across tasks are indicated by asterisks and the bold notation of the gesture or grasping task caption. El-Flex/Ext elbow flexion/extension, FA-Pro/Sup forearm pronation/supination, Mov Time movement time, No of Vel Peak number of peak velocity, Sh-Abd/Add shoulder abduction/adduction, Sh-Flex/Ext shoulder flexion/extension, SPARC spectral arc length, TrunkDisp trunk displacement, Wr-Flex/Ext wrist flexion/extension

Statistical testing (Kruskal–Wallis test) of differences in metric expressions across different tasks of a movement subphase type resulted in statistically significant differences in metric expressions for the majority of conditions. As indicated with asterisks and the bold format of the gesture and grasping task subheadings in Fig. 5A, B, comparable metric expressions were found for trunk displacement in reaching to grasp distally (p = 0.235) and reaching to transport proximally (p = 0.413). Shoulder flexion/extension (p = 0.132) and shoulder abduction/adduction (p = 0.093) were found to be comparable in reaching to grasp distally and reach to gesture distally (shoulder flexion/extension, p = 0.613; shoulder abduction/adduction, p = 0.104). Elbow flexion/extension (p = 0.363) and forearm pronation/supination (p = 0.113) showed similar expressions in proximal reach gesture. Wrist flexion/extension was significantly different across tasks in all four movement subphase conditions. The temporal kinematics, movement time, peak velocity, and SPARC, were comparable between gestures distally as well as proximally with p-values ranging between p = 0.0866 and p = 0.290. Additionally, consistent results across tasks were found for peak velocity in reaching to grasp distally (p = 0.108) and for the NVP during reaching to gesture proximally (p = 0.127).

The relationship between the ten core set kinematic parameters during subphases of reaching distally and proximally were investigated in a correlation matrix for each task, and impairment subgroup. Figure 6 illustrates the strength of correlations for the no, mild and moderate impairment group in a heatmap for reaching distally in blue color code (Fig. 6A) and reaching proximally in green color code (Fig. 6B). Correlation coefficients of r ≥ 0.5, or r ≤ − 0.5 were set as cut-offs to determine significant associations between metrics. Significant associations are highlighted with black square outlines. Metric associations that were expected to be consistent across tasks but that failed to be significantly strong correlated are highlighted with dashed black square outlines.

Fig. 6
figure 6figure 6

A Relationship between kinematics of reaching distally across subjects. Core set kinematics of reaching distally are correlated with each other per movement task and impairment level subgroups. 1, Trunk displacement; 2, Shoulder flexion/extension; 3, Shoulder abduction/adduction; 4, Elbow flexion/extension; 5, Forearm pronation/supination; 6, wrist flexion/extension; 7, Movement time; 8, Peak velocity; 9, Number of velocity peaks (NVP); 10, Spectral arc length (SPARC). The correlation coefficient is presented in a color code as shown on the right of each heat map. Strong correlations between metrics that were consistent across tasks are highlighted by black square outlines in the heatmap. The dashed black square outlines represent metric associations, suspected to be consistent across tasks, that were not significantly strong correlated. B Relationship between kinematics of reaching distally across subjects. Core set kinematics of reaching proximally are correlated with each other per movement task and impairment level subgroups. 1, Trunk displacement; 2, Shoulder flexion/extension; 3, Shoulder abduction/adduction; 4, Elbow flexion/extension; 5, Forearm pronation/supination; 6, wrist flexion/extension; 7, Movement time; 8, Peak Velocity; 9, Number of velocity peaks (NVP); 10, Spectral arc length (SPARC). The correlation coefficient is presented in a color code as shown on the right of each heat map. Strong correlations between metrics that were consistent across tasks are highlighted by black square outlines in the heatmap. The dashed black square outlines represent metric associations, suspected to be consistent across tasks, that were not significantly strong correlated

During reaching distally, differences in terms of shoulder-elbow associations between the no impairment group and the mild and the moderate impairment group were found. Strong associations between elbow flexion/extension and shoulder abduction/adduction are observed across tasks in the no impairment group, whereas predominantly strong associations between elbow and shoulder flexion/extension are visible in the mild and the moderate impairment group. Across impairment groups, the following metric pairs were found to be strongly associated in most tasks; wrist flexion/extension and shoulder flexion/extension, peak velocity and elbow flexion/extension, peak velocity and shoulder flexion/extension, number of velocity peaks and movement time, and number of velocity peaks and spectral arc length.

During reaching proximally, stable relations across tasks were found between peak velocity and shoulder flexion/extension, between the number of velocity peaks and elbow flexion/extension, between the number of velocity peaks and movement time as well as between the NVP and the SPARC. A slight trend towards more consistent correlations is observed across tasks in the movement subgroups of reaching distally and proximally in the moderate impairment group.

Discussion

This study aims to characterize upper limb movement behavior in subjects with and without stroke-related upper limb impairments when performing different activities including gestures and object-use that are associated with everyday life. Based on the presented results, task- and impairment specific kinematic expressions can be confirmed for the whole movement task execution and subphases of reaching distally and proximally that are summarized in the key findings of Box 1.

Box 1 Key findings on upper limb kinematic characteristics

These study findings underpin the importance to consider task specificities in qualitative upper limb movement analysis, as already suggested in the pioneering works of Marc Jeannerod 50 years ago who found that the acceleration phase during reach-to-grasp is shorter compared to the deceleration phase in approach of the grasp, whereas during pointing movements the acceleration phase is considerably longer than the deceleration phase [55]. Furthermore, research suggests that movement tasks, such as rhythmic or discrete tasks, are controlled by different mechanisms [56, 57]. The presented findings on spatiotemporal movement characteristics of moderate to mildly affected stroke subjects in terms of increased trunk compensation and decreased peak velocity are in line with other research on movement kinematics of ADL movements, such as the drinking task assessment [51]. Effects of the task content and difficulty on kinematic outcome measures were detected across impairment severities, in the less-affected upper limb, as well as in healthy subjects [58, 59]. These findings support the consideration of the task difficulty in kinematic measurements is important to minimize floor and ceiling effects of the assessments. Thus, including similar gesture and grasping movements amongst other conditions in standardized assessment protocols would provide further insights into upper limb movement behavior.

Ten kinematic parameters, acquired and processed with a wearable sensing suit, were shown to be useful for assessing the spatiotemporal aspects of upper limb movement behavior for the total task performance as well as for the tasks subphases. The NVP showed good discriminability between different levels of upper limb impairment severity and different tasks. However, it has been criticized for its high task-dependency, in comparison to other smoothness measures, such as SPARC [46]. The strong and consistent relation between the number of velocity peaks and spectral arc length across movement subphases found in the present study, supports the assumption that both measures reflect the movement construct of smoothness. Other studies investigating smoothness measures suggested that normalized jerks are used as a unitless measure that normalizes for both amplitude and duration of the movement [60]. Log dimensionless jerk has been recommended for daily living recordings of different trial durations [61]. Although not explicitly addressed in the present study, the findings support the need to further explore the associations between different smoothness measures in relation to different populations and movement types that are representative for daily living.

Another topic, addressed herein, was the comparability of the metric expression across different tasks on the level of movement subphases. Even though preliminary, some trends and consistencies in movement characteristics of reaching distally and reaching proximally are described across different task types and impairment subgroups. These findings support the assumption of stereotypical motion primitives or building blocks as basic elements of functional motor tasks [35] and could potentially facilitate across task and study comparisons of upper limb kinematic measurements. Schambra et al. found that the two primitives or movement subphases of transport and reach were differentiable based on an unbiased machine learning algorithm with an accuracy of 92.1%. The algorithm’s nodes of the binary tree indicated greater wrist extension, wrist supination and elbow extension in reach, compared to less wrist extension and supination and more shoulder flexion and abduction during transport [34], which is comparable to the results of context-related differences in kinematic expressions in the present study. The algorithm was applied for phase segmentation was based on changes in the hand-sensor position and velocity in x- and z-direction. Visual inspection post-verification determined that the algorithm correctly carried out the segmentation in 85% of the included trials. Failures in automatic phase detection occurred frequently in datasets of subjects with more pronounced deficits that had irregular and slow movement profiles. We therefore conclude that the algorithm reliably detects movement sub phases of reaching to maximum arm length and the head or mouth. The reliable and time-saving segmentation and detection of movement sub phases in this controlled task set is an important prerequisite for the development of future real-life recording and analysis of upper limb functioning after stroke. Automatized movement phase segmentation and analysis has the potential to be useful for real-time feedback during training as well as for developments in assistive devices and training technology. The development of a hand orthosis is an example of the application of task-specific detection of upper limb movement kinematics. In this case, reaching movements with and without the intention to grasp by means of hand and finger sensors to develop the algorithm for controlling the hand orthosis [62]. Other examples of the application are found for specific training of the shoulder muscles, especially the rotator cuff [63] or assistive devices for arm weight support [4, 9, 10]. With respect to the stroke population, the development of devices improving the functionality and quality of upper limb movements after stroke is of great interest.

Some limitations need to be considered when interpreting the results. The number of healthy control subjects was small but well aligned with the stroke sample in terms of age distribution. The fact that this dataset [54] is part of a larger kinematic and kinetic datasets of similar experimental protocols [64] may help to gain a normative dataset on upper limb kinematics. We have not reported about the factor of hand dominance in the single task and subphase analysis. However, the factor of the affected is the dominant-side has been included in the linear-mixed model on gestures vs. grasping movement, without revealing significant effects on any of the tested kinematic parameters. Another limitation could be attributed to the lack of validation reference system for the sensor-based acquired kinematic measures. A systematic review on sensor-based joint angle estimations found measurement errors between 0.71° and 12.1° in the shoulder joint in comparison to optical systems with measurement errors around 1° [65]. Even so, sensor-based measurements were selected for recording the experimental protocol including object interactions and wide workspace contributions to circumvent problems of marker occlusion in optical systems and pave towards applications in flexible environments. Nevertheless, additional measures for movement detection, such as such as video recordings or sensors for interaction force detection, would be recommendable in future research to increase the accuracy of the automatic phase detection algorithm, thereby reducing the time and effort of manual data postprocessing.

Conclusion

In conclusion, the analysis of gesture and grasp movements in a set of activities related to daily living revealed task-specific and impairment-specific characteristics in terms of different kinematic expressions. Grasping motions were characterized by more distally pronounced and slower motions that were less smooth and executed with larger trunk motions in stroke subjects when compared to faster gesture movement, that were shown to be less discriminative between impairment levels. We demonstrated that kinematic assessments of activities of daily living provide general and granular information on movement quality of relevant and natural upper limb motor activities. The comparison of subphases of reaching types across tasks revealed similarities in trunk compensation, shoulder motions and speed as well as smoothness-related metrics. These subphases additionally proved to be differently expressed between persons with no, mild and moderate stroke-related upper limb impairments.

Availability of data and materials

The datasets generated and used for the analysis presented in the current study are available in the Zenodo repository at: https://zenodo.org/record/3713449 [53].

Abbreviations

Abd/add:

Abduction/adduction

BMI:

Body mass index

EmNSA:

Erasmus modified version of the Nottingham Sensory Assessment

Flex/ext:

Flexion/extension

FMMA-UE:

Fugl-Meyer Motor Assessment of the Upper Extremity

MAS:

Modified Ashworth Scale

MoCA:

Montreal Cognitive Assessment

NIHSS:

National Institutes of Health Stroke Scale

NVP:

Number of velocity peaks

Pro/sup:

Pronation/supination

SPARC:

Spectral arc length

References

  1. 1.

    Raghavan P. Upper limb motor impairment after stroke. Phys Med Rehabil Clin N Am. 2015;26(4):599–610. https://doi.org/10.1016/j.pmr.2015.06.008.

    Article  PubMed  PubMed Central  Google Scholar 

  2. 2.

    Cortes JC, Goldsmith J, Harran MD, Xu J, Kim N, Schambra HM, Luft AR, Celnik P, Krakauer JW, Kitago T. A short and distinct time window for recovery of arm motor control early after stroke revealed with a global measure of trajectory kinematics. Neurorehabil Neural Repair. 2017;31(6):552–60. https://doi.org/10.1177/1545968317697034.

    Article  PubMed  PubMed Central  Google Scholar 

  3. 3.

    Bang DH, Shin WS, Choi SJ. The effects of modified constraint-induced movement therapy combined with trunk restraint in subacute stroke: a double-blinded randomized controlled trial. Clin Rehabil. 2015;29:561–9. https://doi.org/10.1177/0269215514552034.

    Article  PubMed  Google Scholar 

  4. 4.

    Bartolo M, De Nunzio AM, Sebastiano F, Spicciato F, Tortola P, Nilsson J, et al. Arm weight support training improves functional motor outcome and movement smoothness after stroke. Funct Neurol. 2014;29(1):15–21.

    PubMed  PubMed Central  Google Scholar 

  5. 5.

    Conrad MO, Gadhoke B, Scheidt RA, Schmit BD. Effect of tendon vibration on hemiparetic arm stability in unstable workspaces. PLoS ONE. 2015;10(12): e0144377. https://doi.org/10.1371/journal.pone.0144377.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  6. 6.

    Corti M, McGuirk TE, Wu SS, Patten C. Differential effects of power training versus functional task practice on compensation and restoration of arm function after stroke. Neurorehabil Neural Repair. 2012;26:842–54. https://doi.org/10.1177/1545968311433426.

    Article  PubMed  Google Scholar 

  7. 7.

    Di Lazzaro V, Capone F, Di Pino G, Pellegrino G, Florio L, Zollo L, et al. Combining robotic training and non-invasive brain stimulation in severe upper limb-impaired chronic stroke patients. Front Neurosci. 2016;10:88. https://doi.org/10.3389/fnins.2016.00088.

    Article  PubMed  PubMed Central  Google Scholar 

  8. 8.

    Durham KF, Sackley CM, Wright CC, Wing AM, Edwards MG, van Vliet P. Attentional focus of feedback for improving performance of reach-to-grasp after stroke: a randomised crossover study. Physiotherapy. 2014;100(2):108–15. https://doi.org/10.1016/j.physio.2013.03.004.

    CAS  Article  PubMed  Google Scholar 

  9. 9.

    Ellis MD, Carmona C, Drogos J, Traxel S, Dewald JP. Progressive abduction loading therapy targeting flexion synergy to regain reaching function in chronic stroke: preliminary results from an rct. Ann Int Con IEEE Eng Med Biol Soc. 2016;2016:5837–40. https://doi.org/10.1109/EMBC.2016.7592055.

    Article  Google Scholar 

  10. 10.

    Housman SJ, Scott KM, Reinkensmeyer DJ. A randomized controlled trial of gravity-supported, computer-enhanced arm exercise for individuals with severe hemiparesis. Neurorehabil Neural Repair. 2009;23:505–14. https://doi.org/10.1177/1545968308331148.

    Article  PubMed  Google Scholar 

  11. 11.

    Hsieh YW, Liing RJ, Lin KC, Wu CY, Liou TH, Lin JC, et al. Sequencing bilateral robot-assisted arm therapy and constraint-induced therapy improves reach to press and trunk kinematics in patients with stroke. J Neuroeng Rehabil. 2016;13:31. https://doi.org/10.1186/s12984-016-0138-5.

    Article  PubMed  PubMed Central  Google Scholar 

  12. 12.

    Kim H, Miller LM, Fedulow I, Simkins M, Abrams GM, Byl N, et al. Kinematic data analysis for post-stroke patients following bilateral versus unilateral rehabilitation with an upper limb wearable robotic system. IEEE Trans Neural Syst Rehabil Eng. 2013;21:153–64. https://doi.org/10.1109/TNSRE.2012.2207462.

    Article  PubMed  Google Scholar 

  13. 13.

    Quattrocchi G, Greenwood R, Rothwell JC, Galea JM, Bestmann S. Reward and punishment enhance motor adaptation in stroke. J Neurol Neurosurg Psychiatry. 2017;88:730–6. https://doi.org/10.1136/jnnp-2016-314728.

    Article  PubMed  Google Scholar 

  14. 14.

    Massie C, Malcolm MP, Greene D, Thaut M. The effects of constraint-induced therapy on kinematic outcomes and compensatory movement patterns: an exploratory study. Arch Phys Med Rehabil. 2009;90:571–9. https://doi.org/10.1016/j.apmr.2008.09.574.

    Article  PubMed  Google Scholar 

  15. 15.

    Wu CY, Chen CL, Tang SF, Lin KC, Huang YY. Kinematic and clinical analyses of upper-extremity movements after constraint-induced movement therapy in patients with stroke: a randomized controlled trial. Arch Phys Med Rehabil. 2007;88:964–70. https://doi.org/10.1016/j.apmr.2007.05.012.

    Article  PubMed  Google Scholar 

  16. 16.

    Wu CY, Chen YA, Chen HC, Lin KC, Yeh IL. Pilot trial of distributed constraint-induced therapy with trunk restraint to improve poststroke reach to grasp and trunk kinematics. Neurorehabil Neural Repair. 2012;26:247–55. https://doi.org/10.1177/1545968311415862.

    Article  PubMed  Google Scholar 

  17. 17.

    Lima RCM, Michaelsen SM, Nascimento LR, Polese JC, Pereira ND, Teixeira-Salmela LF. Addition of trunk restraint to home-based modified constraint-induced movement therapy does not bring additional benefits in chronic stroke individuals with mild and moderate upper limb impairments: a pilot randomized controlled trial. NeuroRehabilitation. 2014;35:391–404. https://doi.org/10.3233/NRE-141130.

    CAS  Article  PubMed  Google Scholar 

  18. 18.

    Colombo R, Pisano F, Delconte C, Mazzone A, Grioni G, Castagna M, et al. Comparison of exercise training effect with different robotic devices for upper limb rehabilitation: a retrospective study. Eur J Phys Rehabil Med. 2017;53:240–8. https://doi.org/10.23736/S1973-9087.16.04297-0.

    Article  PubMed  Google Scholar 

  19. 19.

    Mazzoleni S, Sale P, Tiboni M, Franceschini M, Carrozza MC, Posteraro F. Upper limb robot-assisted therapy in chronic and subacute stroke patients. Am J Phys Med Rehabil. 2013;2013:e26-37. https://doi.org/10.1097/PHM.0b013e3182a1e852.

    Article  Google Scholar 

  20. 20.

    Wu CY, Yang CL, Chuang LL, Lin KC, Chen HC, Chen MD, et al. Effect of therapist-based versus robot-assisted bilateral arm training on motor control, functional performance, and quality of life after chronic stroke: a clinical trial. Phys Ther. 2012;92:1006–16. https://doi.org/10.2522/ptj.20110282.

    Article  PubMed  Google Scholar 

  21. 21.

    Piron L, Turolla A, Agostini M, Zucconi CS, Ventura L, Tonin P, et al. Motor learning principles for rehabilitation: a pilot randomized controlled study in poststroke patients. Neurorehabil Neural Repair. 2010;24:501–8. https://doi.org/10.1177/1545968310362672.

    Article  PubMed  Google Scholar 

  22. 22.

    Subramanian SK, Lourenco CB, Chilingaryan G, Sveistrup H, Levin MF. Arm motor recovery using a virtual reality intervention in chronic stroke: randomized control trial. Neurorehabil Neural Repair. 2013;27:13–23. https://doi.org/10.1177/1545968312449695.

    Article  PubMed  Google Scholar 

  23. 23.

    Lin KC, Chen YA, Chen CL, Wu CY, Chang YF. The effects of bilateral arm training on motor control and functional performance in chronic stroke: a randomized controlled study. Neurorehabil Neural Repair. 2010;24:42–51. https://doi.org/10.1177/1545968309345268.

    Article  PubMed  Google Scholar 

  24. 24.

    Van Delden AEQ, Beek PJ, Roerdink M, Kwakkel G, Peper CE. Unilateral and bilateral upper-limb training interventions after stroke have similar effects on bimanual coupling strength. Neurorehabil Neural Repair. 2015;29:255–67. https://doi.org/10.1177/1545968314543498.

    Article  PubMed  Google Scholar 

  25. 25.

    Wu CY, Yang CL, Chen MD, Lin KC, Wu LL. Unilateral versus bilateral robot-assisted rehabilitation on arm-trunk control and functions post stroke: a randomized controlled trial. J Neuroeng Rehabil. 2013;10:35. https://doi.org/10.1186/1743-0003-10-35.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  26. 26.

    Bensmail D, Robertson JVG, Fermanian C, Roby-Brami A. Botulinum toxin to treat upper-limb spasticity in hemiparetic patients: analysis of function and kinematics of reaching movements. Neurorehabil Neural Repair. 2010;24:273–81. https://doi.org/10.1177/1545968309347683.

    Article  PubMed  Google Scholar 

  27. 27.

    Wu C-Y, Huang P-C, Chen Y-T, Lin K-C, Yang H-W. Effects of mirror therapy on motor and sensory recovery in chronic stroke: a randomized controlled trial. Arch Phys Med Rehabil. 2013;94:1023–30. https://doi.org/10.1016/j.apmr.2013.02.007.

    Article  PubMed  Google Scholar 

  28. 28.

    van Meulen FB, Klaassen B, Held J, Reenalda J, Buurke JH, van Beijnum BF, Luft A, Veltink PH. Objective evaluation of the quality of movement in daily life after stroke. Front Bioeng Biotechnol. 2016;3:210. https://doi.org/10.3389/fbioe.2015.00210.

    Article  PubMed  PubMed Central  Google Scholar 

  29. 29.

    Kwakkel G, Lannin NA, Borschmann K, et al. Standardized measurement of sensorimotor recovery in stroke trials: consensus-based core recommendations from the stroke recovery and rehabilitation roundtable. Int J Stroke. 2017;12(5):451–61. https://doi.org/10.1177/1747493017711813.

    Article  PubMed  Google Scholar 

  30. 30.

    Schwarz A, Kanzler CM, Lambercy O, Luft AR, Veerbeek JM. Systematic review on kinematic assessments of upper limb movements after stroke. Stroke. 2019;50(3):718–27. https://doi.org/10.1161/STROKEAHA.118.023531.

    Article  PubMed  Google Scholar 

  31. 31.

    McCrea PH, Eng JJ, Hodgson AJ. Biomechanics of reaching: clinical implications for individuals with acquired brain injury. Disabil Rehabil. 2002;24(10):534–41. https://doi.org/10.1080/09638280110115393.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  32. 32.

    Alt MM, Hager CK. Kinematic analysis of the upper extremity after stroke—how far have we reached and what have we grasped? Phys Thera Rev. 2015;20(3):137–55. https://doi.org/10.1179/1743288X15Y.0000000002.

    Article  Google Scholar 

  33. 33.

    Alt Murphy M, Murphy S, Persson HC, Bergström UB, Sunnerhagen KS. Kinematic analysis using 3D motion capture of drinking task in people with and without upper-extremity impairments. J Vis Exp. 2018;133:e57228. https://doi.org/10.3791/57228.

    Article  Google Scholar 

  34. 34.

    Schambra HM, Parnandi A, Pandit NG, Uddin J, Wirtanen A, Nilsen DM. A taxonomy of functional upper extremity motion. Front Neurol. 2019;10:857. https://doi.org/10.3389/fneur.2019.00857.

    Article  PubMed  PubMed Central  Google Scholar 

  35. 35.

    Giszter SF. Motor primitives—new data and future questions. Curr Opin Neurobiol. 2015;33:156–65. https://doi.org/10.1016/j.conb.2015.04.004.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  36. 36.

    Averta G, Della Santina E, Battaglia E, Felici F, Bianchi M, Bicchi A. Unvealing the principal modes of human upper limb movements through functional analysis. Front Roboti AI. 2017;4:37. https://doi.org/10.3389/frobt.2017.00037.

    Article  Google Scholar 

  37. 37.

    Schwarz A, Averta G, Veerbeek JM, Luft AR, Held JPO, Valenza G, Bicchi A, Bianchi M. A functional analysis-based approach to quantify upper limb impairment level in chronic stroke patients: a pilot study. Annu Int Conf IEEE Eng Med Biol Soc. 2019;2019:4198–204. https://doi.org/10.1109/EMBC.2019.8857732.

    Article  PubMed  Google Scholar 

  38. 38.

    Feix T, Romero J, Schmiedmayer H-B, Dollar AM, Kragic D. The grasp taxonomy of human grasp types. IEEE Trans Human Mach Syst. 2016;46(1):66–77. https://doi.org/10.1109/THMS.2015.2470657.

    Article  Google Scholar 

  39. 39.

    Roetenberg D, Baten CT, Veltink PH. Estimating body segment orientation by applying inertial and magnetic sensing near ferromagnetic materials. IEEE Trans Neural Syst Rehabil Eng. 2007;15(3):469–71. https://doi.org/10.1109/TNSRE.2007.903946.

    Article  PubMed  Google Scholar 

  40. 40.

    van Meulen FB, Reenalda J, Buurke JH, Veltink PH. Assessment of daily-life reaching performance after stroke. Ann Biomed Eng. 2015;43:478–86. https://doi.org/10.1109/70.34763.

    Article  PubMed  Google Scholar 

  41. 41.

    Klaassen B, van Beijnum BF, Held JP, Reenalda J, van Meulen FB, Veltink PH, Hermens HJ. Usability evaluations of a wearable inertial sensing system and quality of movement metrics for stroke survivors by care professionals. Front Bioeng Biotechnol. 2017;5:20. https://doi.org/10.3389/fbioe.2017.00020.

    Article  PubMed  PubMed Central  Google Scholar 

  42. 42.

    Held JPO, Klaassen B, Eenhoorn A, van Beijnum BF, Buurke JH, Veltink PH, Luft AR. Inertial sensor measurements of upper-limb kinematics in stroke patients in clinic and home environment. Front Bioeng Biotechnol. 2018;6:27. https://doi.org/10.3389/fbioe.2018.00027.

    Article  PubMed  PubMed Central  Google Scholar 

  43. 43.

    Höglund G, Grip H, Öhberg F. The importance of inertial measurement unit placement in assessing upper limb motion. Med Eng Phys. 2021;92:1–9. https://doi.org/10.1016/j.medengphy.2021.03.010.

    Article  PubMed  Google Scholar 

  44. 44.

    Xsens MVN User Manual, User Guide Xsens MVN, MVN Link, MVN Awinda, Document 760 MV0319P, Revision X, Oct 2018 Xsens, Enschede, NL, Accessed 10 Nov 2018.

  45. 45.

    Subramanian SK, Yamanaka J, Chilingaryan G, Levin MF. Validity of movement pattern kinematics as measures of arm motor impairment poststroke. Stroke. 2010;41:2303–8. https://doi.org/10.1161/STROKEAHA.110.593368.

    Article  PubMed  Google Scholar 

  46. 46.

    Balasubramanian S, Melendez-Calderon A, Roby-Brami A, Burdet E. On the analysis of movement smoothness. J Neuroeng Rehabil. 2015;12(1):112. https://doi.org/10.1186/s12984-015-0090-9.

    Article  PubMed  PubMed Central  Google Scholar 

  47. 47.

    Repnik E, Puh U, Goljar N, Munih M, Mihelj M. Using inertial measurement units and electromyography to quantify movement during action research arm test execution. Sensors (Basel). 2018;18(9):2767. https://doi.org/10.3390/s18092767.

    Article  Google Scholar 

  48. 48.

    Schwarz A, Bhagubai MMC, Wolterink G, Held JPO, Luft AR, Veltink PH. Assessment of upper limb movement impairments after stroke using wearable inertial sensing. Sensors (Basel). 2020;20(17):4770. https://doi.org/10.3390/s20174770.

    Article  Google Scholar 

  49. 49.

    Ortega-Bastidas P, Aqueveque P, Gómez B, Saavedra F, Cano-de-la-Cuerda R. Use of a single wireless IMU for the segmentation and automatic analysis of activities performed in the 3-m timed up & go test. Sensors (Basel). 2019;19(7):1647. https://doi.org/10.3390/s19071647.

    Article  Google Scholar 

  50. 50.

    Fugl-Meyer AR, Jääskö L, Leyman I, Olsson S, Steglind S. The post-stroke hemiplegic patient. 1. A method for evaluation of physical performance. Scan J Rehabil Med. 1975;7:13–31.

    CAS  Google Scholar 

  51. 51.

    Alt MM, Willen C, Sunnerhagen KS. Kinematic variables quantifying upper-extremity performance after stroke during reaching and drinking from a glass. Neurorehabil Neural Repair. 2011;25:71–80. https://doi.org/10.1177/1545968310370748.

    Article  Google Scholar 

  52. 52.

    Persson HC, Alt Murphy M, Danielsson A, Lundgren-Nilsson Å, Sunnerhagen KS. A cohort study investigating a simple, early assessment to predict upper extremity function after stroke—a part of the SALGOT study. BMC Neurol. 2015;15:92. https://doi.org/10.1186/s12883-015-0349-6.

    Article  PubMed  PubMed Central  Google Scholar 

  53. 53.

    Hoonhorst MH, Nijland RH, van den Berg JS, Emmelot CH, Kollen BJ, Kwakkel G. How do fugl-meyer arm motor scores relate to dexterity according to the action research arm test at 6 months poststroke? Arch Phys Med Rehabil. 2015;96(10):1845–9. https://doi.org/10.1016/j.apmr.2015.06.009.

    Article  PubMed  Google Scholar 

  54. 54.

    Schwarz A, Held, JPO, Luft, AR. Post-stroke upper limb kinematics of a set of daily living tasks (version 1.0.0) . Zenodo. 2020; https://doi.org/10.5281/zenodo.3713449.

  55. 55.

    Jeannerod M. The neural and behavioural organization of goal-directed movements. Oxford: Clarendon Press; 1988.

    Google Scholar 

  56. 56.

    Hogan N, Sternad D. On rhythmic and discrete movements: reflections, definitions and implications for motor control. Exp Brain Res. 2007;181(1):13–30. https://doi.org/10.1007/s00221-007-0899-y.

    Article  PubMed  Google Scholar 

  57. 57.

    Huys R, Studenka BE, Rheaume NL, Zelaznik HN, Jirsa VK. Distinct timing mechanisms produce discrete and continuous movements. PLoS Comput Biol. 2008;4(4): e1000061. https://doi.org/10.1371/journal.pcbi.1000061.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  58. 58.

    Hussain N, Alt Murphy M, Sunnerhagen KS. Upper limb kinematics in stroke and healthy controls using target-to-target task in virtual reality. Front Neurol. 2018;9:300. https://doi.org/10.3389/fneur.2018.00300.

    Article  PubMed  PubMed Central  Google Scholar 

  59. 59.

    Poston B, Van Gemmert AW, Barduson B, Stelmach GE. Movement structure in young and elderly adults during goal-directed movements of the left and right arm. Brain Cogn. 2009;69:30–8. https://doi.org/10.1016/j.bandc.2008.05.002.

    Article  PubMed  Google Scholar 

  60. 60.

    Buma FE, van Kordelaar J, Raemaekers M, van Wegen EEH, Ramsey NF, Kwakkel G. Brain activation is related to smoothness of upper limb movements after stroke. Exp Brain Res. 2016;234(7):2077–89. https://doi.org/10.1007/s00221-015-4538-8.

    Article  PubMed  PubMed Central  Google Scholar 

  61. 61.

    Gulde P, Hermsdörfer J. Smoothness metrics in complex movement tasks. Front Neurol. 2018;9:615. https://doi.org/10.3389/fneur.2018.00615.

    Article  PubMed  PubMed Central  Google Scholar 

  62. 62.

    de Vries JC, van Ommeren AL, Prange-Lasonder GP, Rietman JS, Veltink PH. Detection of the intention to grasp during reach movements. J Rehabil Assist Technol Eng. 2018;5:2055668317752850. https://doi.org/10.1177/2055668317752850.

    Article  PubMed  PubMed Central  Google Scholar 

  63. 63.

    Janssen MMHP, Horstik J, Klap P, de Groot IJM. Feasibility and effectiveness of a novel dynamic arm support in persons with spinal muscular atrophy and duchenne muscular dystrophy. J Neuroeng Rehabil. 2021;18(1):84. https://doi.org/10.1186/s12984-021-00868-6.

    Article  PubMed  PubMed Central  Google Scholar 

  64. 64.

    Averta G, Barontini F, Catrambone V, Haddadin S, Handjaras G, Held JPO, et al. U-Limb: a multi-modal, multi-center database on arm motion control in healthy and post-stroke conditions. GigaScience. 2021;10(6):giab043. https://doi.org/10.1093/gigascience/giab043.

    Article  PubMed  PubMed Central  Google Scholar 

  65. 65.

    Garimella R, Peeters T, Beyers K, Truijen S, Huysmans T, Verwulgen S. Capturing joint angles of the off-site human body. IEEE Sens. 2018;2018:1–4. https://doi.org/10.1109/ICSENS.2018.8589711.

    Article  Google Scholar 

Download references

Acknowledgements

We thank all participants for their attendance in the study.

Funding

This project has received funding from the European Union’s Horizon 2020 Research and Innovation Programme under grant agreement number 688857 (SoftPro), from the Swiss State Secretariat for Education, Research and Innovation (contract number 15.0283-1), and the P&K Pühringer Foundation.

Author information

Affiliations

Authors

Contributions

AS, JH, and AL designed and planned the study. AS and SHGN were involved in data acquisition and performed the data curation and analysis in collaboration with MMCB. AS took lead in writing the manuscript. JHB and PHV reviewed the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Anne Schwarz.

Ethics declarations

Ethics approval and consent to participate

The study protocol of this observational was approved by the Cantonal Ethics Committee Northwest and Central Switzerland (BASEC-ID: 2016-02075), prospectively registered in ClinicalTrials.gov (NCT03135093). All participants gave informed consent before inclusion.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Schwarz, A., Bhagubai, M.M.C., Nies, S.H.G. et al. Characterization of stroke-related upper limb motor impairments across various upper limb activities by use of kinematic core set measures. J NeuroEngineering Rehabil 19, 2 (2022). https://doi.org/10.1186/s12984-021-00979-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12984-021-00979-0

Keywords

  • Biomechanical phenomena
  • Kinematics
  • Upper extremity
  • Stroke