Figure 1 - uploaded by Nestor Jarque-Bou
Content may be subject to copyright.
Experiment A: List of activities of daily living (ADLs) performed and scenario with the objects used. Unless indicated otherwise, the position of the subject was standing.
Source publication
Simultaneous measurement of the kinematics of all hand segments is cumbersome due to sensor placement constraints, occlusions, and environmental disturbances. The aim of this study is to reduce the number of sensors required by using kinematic synergies, which are considered the basic building blocks underlying hand motions. Synergies were identifi...
Contexts in source publication
Context 1
... criteria used to select subjects were gender parity in the overall data, age between 20 and 65 years, and no declared upper limb pathologies. Subjects performed 26 representative ADL (Figure 1): 20 activities adapted from the Sollerman Hand Function Test (to ensure their repeatability and to favor their standardization), and six additional activities (A10, A15, A19, A24, A25, A26) that were added based on the percentage of use of the commonest grasps in ADL [42]. In order to foster repeatability, precise instructions for each task were provided and each ADL started and ended with the body and arms in the same posture (arms and hands relaxed at the side of the body when subjects were standing, or arms and hands resting in a relaxed position on the table when they were sitting). ...
Context 2
... an example, Figures 9 and 10 show the temporal evolution of the estimated angle versus the recorded angle for subject 2 during two actions from Experiment B: the one with the highest RMSE on average (Figure 9) and the one with the lowest error ( Figure 10). Figure 8 shows the global RMSE errors, expressed as a percentage with respect to the RoM of each DoF, per action and joint (computed from all frames and subjects of each action) for both Experiment A (A1-A26) and Experiment B (B1-B33). ...
Context 3
... an example, Figures 9 and 10 show the temporal evolution of the estimated angle versus the recorded angle for subject 2 during two actions from Experiment B: the one with the highest RMSE on average (Figure 9) and the one with the lowest error ( Figure 10). Figure 8 shows the global RMSE errors, expressed as a percentage with respect to the RoM of each DoF, per action and joint (computed from all frames and subjects of each action) for both Experiment A (A1-A26) and Experiment B (B1-B33). ...
Context 4
... baking powder into the bowl" (AVG = 6.1%). As an example, Figures 9 and 10 show the temporal evolution of the estimated angle versus the recorded angle for subject 2 during two actions from Experiment B: the one with the highest RMSE on average (Figure 9) and the one with the lowest error ( Figure 10). ...
Context 5
... baking powder into the bowl" (AVG = 6.1%). As an example, Figures 9 and 10 show the temporal evolution of the estimated angle versus the recorded angle for subject 2 during two actions from Experiment B: the one with the highest RMSE on average (Figure 9) and the one with the lowest error ( Figure 10). ...
Context 6
... have provided a detailed description of the most independent hand DoF as well as the most coordinated ones during the performance of representative ADL. Furthermore, we pro- Figure 10. Estimated versus recorded angles for action 21 ("Pouring baking powder into the bowl"). 1 to 5 refers to thumb to little fingers. ...
Context 7
... particular, for each of the 21 combinations tested (Table 3), the best solution, in terms of lower average RMSE across DoF, was the combination that used flexion of MCP and PIP joints of the ring finger ( Figure 11). This result seems quite reasonable, since the thumb and index finger are more independent, and the one that best approaches the others is the ring, which is more centered with the other two fingers. ...
Citations
... Additionally, the concept of an across-subject calibration [11] is considered, as it reduces the number of measurements required for the calibration with each person, making it suitable for large clinical studies or even regular clinical routine. Another approach worth mentioning is the concept of dimensionality reduction, i. e., kinematic synergies that establish a relation between joint angles [22]. This approach has to be used with care, as it deduces all joint angles by measuring only reduced number after kinematic synergies were found. ...
... We did not investigate the wrist sensors and the palm arch has shown to be difficult to calibrate. Even though the findings in this work need to be confirmed for arthritis patients as well, it already shows that using the CyberGlove III for objective hand monitoring of healthy subjects is feasible, supporting previous findings ( [2], [6], [8], [11], [12], [22]). Furthermore, the CyberGlove III was already used to discriminate between healthy individual and patients suffering from osteoarthritis with an accuracy of 80 % [8]. ...
We investigate the accuracy of different calibration methods for the CyberGlove III for objective hand function assessments. The accuracy is evaluated by means of root-mean-squared errors (RMSE) between ground truth and angles estimated by the glove. Additionally, we propose two improvements. The first increases standardization capabilities for the measurement of pure thumb carpometacarpal joint flexion. The second increases the accuracy by extending an existing calibration method to all thumb and distal interphalangeal joints. The best calibration method is identified and compared to an across-subject calibration by means of RMSEs, highlighting the trade-off between the number of necessary measurements and the accuracy. Our proposed improvements both reduce the measurement effort, while only the second improvement reduces the RMSEs substantially. The best calibration method yields RMSEs below 10° for most joints, while the across-subject calibration has RMSEs that are approximately 5° to 10° higher. We conclude that the CyberGlove III is a suitable tool for objective monitoring of hand movement and function, and the across-subject calibration has the potential to track the range of motion or frequency of joint movements. With this work, we provide an overview for researchers to choose calibration methods suitable for their application depending on necessary accuracy and possible extent of calibration measurements. This work further highlights potential obstacles, when such a glove is used with patients.
... This method looks for linear combinations of correlated variables to find a reduced set of new uncorrelated variables. In previous studies of the authors [18,20], kinematic synergies have been applied to reconstruct the full hand kinematics by recording only a few joint angles and estimating the remaining angles from the coordination established by those synergies. This has been verified to be feasible in not only healthy subjects [20], but also in patients [18]. ...
... In previous studies of the authors [18,20], kinematic synergies have been applied to reconstruct the full hand kinematics by recording only a few joint angles and estimating the remaining angles from the coordination established by those synergies. This has been verified to be feasible in not only healthy subjects [20], but also in patients [18]. ...
... The reduction of tasks was performed efficiently by comparing subject-specific hand kinematics in terms of synergies (first four synergies), instead of comparing the original 16 joint angles. The kinematic synergies herein obtained per subject after considering all the SHFT tasks were similar to those reported in the literature [20,27,29,31]: the first two synergies were related to fingers flexion and adduction, and were similar among subjects. The third and fourth synergies showed independence of thumb [28,32], and were more diverse between subjects. ...
Background
Hand kinematics during hand function tests based on the performance of activities of daily living (ADLs) can provide objective data to determine patients’ functional loss. However, they are rarely used during clinical assessments because of their long duration. Starting with the 20 Sollerman Hand Function Test (SHFT) tasks, we propose identifying a reduced set of ADLs that provides similar kinematic information to the original full set in terms of synergies, ranges of motion and velocities.
Methods
We followed an iterative method with the kinematics of 16 hand joints while performing the 20 ADLs of the SHFT. For each subject, ADLs were ordered according to their influence on the synergies obtained by means of a principal component analysis, the minimum number of ADLs that represented the original kinematic synergies (maximum angle of 30° between synergies), and the maintained ranges of joint movements (85% of the original ones) were selected for each subject. The set of the most frequently selected ADLs was verified to be representative of the SHFT ADLs in terms of motion strategies, ranges of motion and joint velocities when considering healthy subjects and Hand Osteoarthritis patients.
Results
A set of 10 tasks, the BE-UJI activity set, was identified by ensuring a certain (minimum) similarity in synergy (maximum mean angle between synergies of 25.5°), functional joint ranges (maximum differences of 10°) and joint velocities (maximum differences of 15°/s). The obtained tasks were: pick up coins from purses, lift wooden cubes, pick up nuts and turn them, write with a pen, cut with a knife, lift a telephone, unscrew jar lids and pour water from a cup, a jar and a Pure-Pak. These activities guarantee using the seven commonest handgrips in ADLs.
Conclusion
The BE-UJI activity set for the hand function assessment can be used to obtain quantitative data in clinics as an alternative to the SHFT. It reduces the test time and allows clinicians to obtain objective kinematic data of the motor strategies, ranges of motion and joint velocities used by patients.
... While such a method would help in identifying joints that portray coordination while explaining maximum variance, the method would be participant-specific and cannot be generalized since the synergies are different for different participants. However, synergies acquired from data recorded during ADL can be clustered to obtain a set of synergies that are representative of the global population [33]. While this approach provides promising results, it is not always possible to obtain a cluster of synergies that would contain synergies from all participants [34]. ...
The human hand needs a large number of sensors to measure kinematics owing to its large number of degrees of freedom. Existing devices like data gloves and optical trackers are associated with calibration, line of sight, and accuracy problems. In this paper, we attempt to measure the full hand kinematics using Electromagnetic Tracking Sensors (EMTS) which are accurate, free of line-of-sight problems, and require no calibration. However, EMTS provides output in the form of rotation groups which are defined on a nonlinear manifold. Hence, linear operations required for experimental analysis such as linear dimensionality reduction are not valid. Also, these sensors are expensive, utilize space in terms of cabling, and require a reduced sensor layout. In this paper, we present measurement methods, test the utility of linear and non-linear dimensionality reduction techniques on quaternions and exponential maps. We also performed sensor reduction using a Gini feature selection based on random forest algorithm. The kinematic measurement results show that EMTS yield superior posture reproduction with an error of less than 1 degree. Autoencoder, a nonlinear dimensionality reduction technique, was successfully applied on quaternions which was tuned to perform better than Principal Component Analysis (PCA) in reducing dimensions. The reduced sensor layout with 8 sensors was able to predict full hand kinematics with a Root Mean Square Error (RMSE) of 5.1 degrees.
... Furthermore, the observed kinematic synergies could also be applied to reconstruct entire hand kinematics from the recording of only a few joint angles by estimating the remaining angles from the coordination established by those synergies. This has already been proven to be feasible in healthy subjects [40] but needs to be studied in patients. Lowering the number of hand joints to be recorded would reduce the investment required. ...
... The errors in estimating the hand joint angles from the measurement of only five joint angles (flexion and abduction of thumb carpometacarpal joint, flexion of metacarpophalangeal joint of the middle finger and proximal interphalangeal joint of the ring finger, and abduction between ring and little fingers) are small (7.0 • ± 2.7 • ). The errors found are of the same order of magnitude as the errors from the recording technique [40]. Despite being only a preliminary exploration, the results obtained suggest the feasibility of acquiring entire hand kinematics from the recording of only a few angles. ...
... Despite being only a preliminary exploration, the results obtained suggest the feasibility of acquiring entire hand kinematics from the recording of only a few angles. Further studies are required to make it clear whether the proposed angles are the most suitable ones or if more sophisticated methods for looking for the most appropriate ones [40] would be needed. Searching for indicators of kinematic dysfunction based only on the recordings of these few angles would, therefore, be interesting. ...
Sensorized gloves allow the measurement of all hand kinematics that are essential for daily functionality. However, they are scarcely used by clinicians, mainly because of the difficulty of analyzing all joint angles simultaneously. This study aims to render this analysis easier in order to enable the applicability of the early detection of hand osteoarthritis (HOA) and the identification of indicators of dysfunction. Dimensional reduction was used to compare kinematics (16 angles) of HOA patients and healthy subjects while performing the tasks of the Sollerman hand function test (SHFT). Five synergies were identified by using principal component (PC) analyses, patients using less fingers arch, higher palm arching, and a more independent thumb abduction. The healthy PCs, explaining 70% of patients’ data variance, were used to transform the set of angles of both samples into five reduced variables (RVs): fingers arch, hand closure, thumb-index pinch, forced thumb opposition, and palmar arching. Significant differences between samples were identified in the ranges of movement of most of the RVs and in the median values of hand closure and thumb opposition. A discriminant function for the detection of HOA, based in RVs, is provided, with a success rate of detection higher than that of the SHFT. The temporal profiles of the RVs in two tasks were also compared, showing their potentiality as dysfunction indicators. Finally, reducing the number of sensors to only one sensor per synergy was explored through a linear regression, resulting in a mean error of 7.0°.
Background. Hand kinematics during hand function tests based on the performance of activities of daily living (ADL) can provide objective data to determine patients’ functional loss. However, they are rarely used during clinical evaluation due to their long duration. Starting from the 20 tasks of the Sollerman Hand Function Test (SHFT), we propose to identify a smaller set of ADLs that provides similar kinematic information to the original full set, in terms of synergies, ranges of motion and velocities.
Methods. We used an iterative method with the kinematics of 16 joints of the hand while performing the 20 ADLs of the SHFT. For each subject, the ADLs were ordered according to their influence on the synergies obtained by means of principal component analysis, and the minimum number of ADL that represented the original kinematic synergies (maximum angle of 30 degrees between synergies) and keeping the ranges of joint movements (85% of the original ones) were selected for each subject. Then, the set of the most frequently selected ADLs was verified to be representative of the SHFT ADLs in terms of motion strategies, ranges of motion and joint velocities, considering healthy subjects and Hand Osteoarthritis patients.
Results. A set of 10 tasks, the BE-UJI activity set, was identified, ensuring a certain bounded synergy similarity (maximum mean angle between synergies of 25.5 degrees), functional joint ranges (maximum differences of 10 degrees) and joint velocities (maximum differences of 15 degrees per second). Tasks obtained are: Pick up coins from purses, lift wooden cubes, pick put nuts and turn them, write with pen, cut with a knife, lift telephone, unscrew lid of jars and pour water from a cup, a jar and a Pure-Pak. These activities warranty the use of the seven most common hand-grips in ADL.
Conclusion. The BE-UJI activity set for the hand function assessment could be used in clinics as an alternative to the SHFT, reducing the time of the assessment while allowing clinicians to obtain objective kinematic data of motor strategies, range of motion and joint velocities used by patients.
