Sharing of hand kinematic synergies across subjects in daily living activities

Abstract

Roda-Sales the motor system is hypothesised to use kinematic synergies to simplify hand control. Recent studies suggest that there is a large set of synergies, sparse in degrees of freedom, shared across subjects, so that each subject performs each action with a sparse combination of synergies. identifying how synergies are shared across subjects can help in prostheses design, in clinical decision-making or in rehabilitation. Subject-specific synergies of healthy subjects performing a wide number of representative daily living activities were obtained through principal component analysis. to make synergies comparable between subjects and tasks, the hand kinematics data were scaled using normative range of motion data. to obtain synergies sparse in degrees of freedom a rotation method that maximizes the sum of the variances of the squared loadings was applied. Resulting synergies were clustered and each cluster was characterized by a core synergy and different indexes (prevalence, relevance for function and within-cluster synergy similarity), substantiating the sparsity of synergies. The first two core synergies represent finger flexion and were present in all subjects. The remaining core synergies represent coordination of the thumb joints, thumb-index joints, palmar arching or fingers adduction, and were employed by subjects in different combinations, thus revealing different subjectspecific strategies.

Synergy-Based Sensor Reduction for Recording the Whole Hand Kinematics

“…First, the method they used undervalued DoF with small RoM: Synergies in PCA using the covariance matrix are quite sensitive to the variances of the original variables so that those with larger ranges of variation dominate over those with small ranges. In order to ensure each variable contributes equally to the analysis, PCA may be applied to the covariance matrix with standardized data (mean = 0 and SD = 1 for each DoF) [ 44 ], as has been carried out in this study. This results in a higher number of synergies identified per subject.…”

“…Second, they considered synergies extracted from only one subject, and were therefore not representative of the overall population. As we have seen, all subjects share the same first two synergies (PIP flexion coordination and MCP flexion/abduction coordination) while the rest of the synergies seem to depend on the different strategies employed by each specific subject [ 43 , 44 ]. Third, the synergies were extracted from non-representative activities (free motion while imagining grasping 57 imaginary objects).…”

“…A reference posture (hand resting flat on a table with fingers and thumbs close together, middle fingers aligned with forearms) was recorded before recording the hand kinematics while performing the selected ADL, and was considered zero for all the joint angles [ 25 ]. The recorded joint angles were filtered by a 2nd-order 2-way low-pass Butterworth filter with a cut-off frequency of 5 Hz [ 43 , 44 ], and the initial and final frames of each recording during which the hands remained static were trimmed.…”

A soft, synergy-based robotic glove for grasping assistance

“…Although counterintuitively, a large fraction of ADLs do not require fine control of each DoF of the hand. There is increasing evidence that many movements used in the large variety of ADLs involving the hand can be deconstructed into a narrow set of coordinated postural synergies (Santello et al, 1998;Della Santina et al, 2017;Gracia-Ibáñez et al, 2020).…”

“…To design for and understand the mechanical complexity of the human hand, with more than 24 DoFs, grasp synergies have become a fundamental solution toward reducing this high dimensionality (Liu et al, 2016;Gracia-Ibáñez et al, 2020). The wide variety of hand postures can be deconstructed into a small number of principal components, the first two of which account for 84% of the variance of ADL grasp postures of common objects (Santello et al, 1998).…”