A calibrated database of kinematics and EMG of the forearm and hand during activities of daily living

Néstor J Jarque-Bou,Alba Roda-Sales,Verónica Gracia-Ibáñez,Margarita Vergara,Joaquín L Sancho-Bru

doi:10.1038/s41597-019-0285-1

Abstract

Linking hand kinematics and forearm muscle activity is a challenging and crucial problem for several domains, such as prosthetics, 3D modelling or rehabilitation. To advance in this relationship between hand kinematics and muscle activity, synchronised and well-defined data are needed. However, currently available datasets are scarce, and the presented tasks and data are often limited. This paper presents the KIN-MUS UJI Dataset that contains 572 recordings with anatomical angles and forearm muscle activity of 22 subjects while performing 26 representative activities of daily living. This dataset is, to our knowledge, the biggest currently available hand kinematics and muscle activity dataset to focus on goal-oriented actions. Data were recorded using a CyberGlove instrumented glove and surface EMG electrodes, both properly synchronised. Eighteen hand anatomical angles were obtained from the glove sensors by a validated calibration procedure. Surface EMG activity was recorded from seven representative forearm areas. The statistics verified that data were not affected by the experimental procedures and were similar to the data acquired under real-life conditions.

Highlights

Background & SummaryThe hand is a complex functional limb with more than 20 joints controlled by more than 30 muscles that allow a wide range of activities to be performed very precisely
Some hand kinematics datasets are available in the literature, as well as forearm EMG datasets, very few datasets exist with simultaneously recorded kinematics and EMG9–11, and they have their weaknesses:
Note that some sensors have non-linear relationships with anatomical angles due to either their position or the influence of other joint movements[14], and require using calibration procedures to obtain reliable angles, like that described in a previous work[15], with a mean precision error of 4.45 degrees

Summary

Introduction

Background & SummaryThe hand is a complex functional limb with more than 20 joints controlled by more than 30 muscles that allow a wide range of activities to be performed very precisely. A previous work[17] has identified the most representative forearm areas (from identifiable landmarks) to improve sEMG electrodes placement for ADL performance in EMG activity terms. The presented KIN-MUS UJI dataset[19] aims to allow worldwide research groups to study the relationship between hand kinematics and muscle activity required to perform ADL.

Results

Conclusion