Classification of power grip and precision grip in children using an EIT-based tactile sensor

Quantitative monitoring and measurement of hand motion in children are crucial to support healthy development. Electrical impedance tomography-based tactile sensors, also known as tomographic tactile sensors, provide a promising approach for grasp classification. Our previous study in adults and children demonstrated the feasibility of pinch classification using a cylindrical device equipped with the tomographic tactile sensor. In this study, we developed a new sensing device to classify the power grip and precision grip in children. In order to address concerns that children might lick or swing the device, a cylindrical sensing device was integrated sensor and measurement circuit, incorporated a protective layer for enhanced safety. Seventeen children participated in an experiment to evaluate the feasibility of the grasp classification. The classification features were voltage vectors and reconstructed images obtained from the sensor, and two machine learning methods were used as the classifiers. The average classification accuracy exceeded 85% for both feature types, surpassing the chance level of 50%. These results demonstrate that the basic grasp patterns in children can be accurately classified using a tomographic tactile sensor. This study provides new insights into the future application of grasp motion classification in children.