Reviewed by Lexie CornerMay 22 2024
Researchers from the Shibaura Institute of Technology have developed a device using a flexible tactile sensor to assess fine finger movements. This device finds application in education and medicine, especially in helping infants’ cognitive development. The research was published in the journal IEEE Access.
Human cognition heavily depends on fine motor skills, which impact everything from everyday tasks to the emergence of highly developed tool-based civilizations. It has been difficult to measure and assess these abilities objectively.
Even if they are effective, conventional methods like video coding take a lot of time and are prone to coder bias. Current methods, such as hand-attached devices or marker-less motion capture, have limitations, particularly when evaluating the finger motions of infants.
A ground-breaking study led by Professor Hiroki Sato of the Shibaura Institute of Technology, in association with Mr. Ryunosuke Asahi of the same institution and Dr. Shunsuke Yoshimoto of the University of Tokyo, which is currently a part of Osaka University, has surfaced in response to these challenges. This study presents a fresh approach to the objective assessment of dexterous finger movements.
This work introduces a state-of-the-art Electrical Impedance Tomography (EIT) based flexible touch sensor system.
We have extended Dr. Yoshimoto's previously developed flexible tactile sensor based on electrical impedance tomography (EIT). This extension has resulted in the creation of a novel system for objective evaluation of fine finger movements. This system offers significant advantages in terms of flexibility, shape versatility, and sensitivity compared to conventional methods.
Hiroki Sato, Professor, Shibaura Institute of Technology
Their system is shaped like a cylindrical FDT pegboard (Functional Dexterity Test). It consists of four layers, one of which is a flexible tactile sensor based on Electrical Impedance Tomography (EIT). With this configuration, pinching motions can be precisely measured.
The sensor recorded voltage data from various finger movements using conductive materials and a flexible printed circuit board with 16 electrodes. MATLAB was used to process the data to classify pinching motions and rebuild images. Using recorded voltage vectors and reconstructed pictures, this system demonstrated good classification accuracy in 12 participant trials.
The research introduces a system that can effectively distinguish different pinching actions, thereby addressing a significant gap in existing evaluation methodologies.
In this study, 12 adult participants performed six types of pinching motions characterized by the number of fingers and their direction. The voltage vector and reconstructed images were used to classify the six types of pinching motions. Our results showed classification accuracies of 79.1 % and 91.4 % for the use of reconstructed images and measured voltage vectors, respectively.
Hiroki Sato, Professor, Shibaura Institute of Technology
This discovery has significant ramifications for both theoretical and applied research. Notably, it can open the door for educational toys that support cognitive growth by improving finger dexterity. Computerized hand movement analysis can also help close the workforce gap in medical research and advance the development of online health services.
In the future, we plan to apply the tomographic tactile sensor to objects of various shapes to confirm its feasibility for a wide range of people, particularly infants.
Hiroki Sato, Professor, Shibaura Institute of Technology
This innovative technique represents a significant breakthrough in the objective assessment of dexterous finger movements. With its potential applications in a wide range of domains, including medical research and developmental evaluations, this technology holds out hope for a better future in which the complexities of human motor skills are better understood and used for the good of society.
Journal Reference:
Asahi, R., et al. (2024) Development of Pinching Motion Classification Method using EIT-based Tactile Sensor. IEEE Access. doi.org/10.1109/access.2024.3395271.