It is important to maintain the sensation of moving our bodies as we wish and feeling various things for ourselves in order to enjoy our daily lives even as we age. Tools and information technology have the potential to overcome the barriers and limitations of our physical abilities and bring us unknown experiences. We are developing human-enhancement assistive devices that extend and improve human movement and sensation functions and capabilities to realize this.

Tactile sensation, one of the components of texture, significantly impacts product evaluation, so a quantitative evaluation method for tactile sensation is needed. Conventional tactile evaluation techniques, however, require the preparation of actual samples to be measured by sensors, which is costly in terms of sample production. To solve this problem, we are developing a Digital Haptics Design (DSD) tool that supports the creation of texture patterns on a computer by predicting the tactile response in advance.

We aim to apply engineering and information processing technologies to different fields and develop devices that support discoveries and specialists. In particular, we focus on applications to the medical, nursing, and healthcare fields and conduct joint research with medical professionals.

Advances in computer technology are making it possible to reproduce the musculoskeletal system's physiological, mechanistic, and mechanical properties using computers. By applying these body modeling techniques, we are researching methods to evaluate the characteristics of human movement and sensation with clear physical evidence and to link this to assistive technologies for smarter human support and more user-friendly product design.

Using 126 electrodes placed on the bottom of an aquarium, we measured the bioelectrical signals of zebrafish (Danio rerio), which are about 3 cm long. By extracting respiratory and motor information from the bioelectrical signals and discriminating them using a neural network, we have succeeded in estimating the emotional states of the fish, such as fear/anxiety and pleasure states.

We measured brain activity, peripheral arterial stiffness as an index of peripheral sympathetic nerve activity, and subjective evaluation using visual analogue scales (VASs) when painful stimuli were applied to the skin surface of the participants. Experimental results clarified that the brain activities in some pain-related brain areas were significantly correlated with the peripheral arterial stiffness.

We analyzed the relationships among changes in vascular dynamics characteristics governed by sympathetic nerve activity, skin electrical stimulation accompanied with pain, and subjective pain intensity. As results, certain relationships were found among them.

We proposed a novel sphygmomanometer that enable an evaluator to palpate the carotid artery wave of the subject. Specifically, the proposed sphygmomanometer has the following advantages: (1) continuous arterial pressure with a high SN ratio can be measured by the proposed sensor; (2) the evaluator can perceive the carotid artery pressure through the sphygmomanometer; and (3) The sphygmomanometer is compact and easy to detach.

Enclosed zone flow-mediated vasodilation (ezFMD) is a vascular endothelial function evaluation method using the oscillometric method. We proposed a novel endothelial function evaluation index considering the mechanical measurement principle of ezFMD, and validated its effectiveness.

In this study, we proposed a log-linearized peripheral vascular viscoelastic index by deriving an approximate model of vascular dynamics that considers the intravascular pressure in the peripheral vasculature and the nonlinear characteristics of the vascular system.

This study proposed a neural network model of Caenorhabditis elegans, for which the connectome has been elucidated. The connectome-based neural network model enables the simulation of chemotaxis, and is thus expected to contribute to the design of guidelines for actual biological experiments that aim to determine the information processing mechanism of chemotaxis.

This study focused on the bioelectric (ventilatory) signals of small fish and proposed a bioassay system for detecting water pollution by simultaneously monitoring fish behavior and ventilatory signals.

To predict the perceptual similarities of odor pairs from neural activities of the olfactory system, we proposed a neural net model based on the olfactory system structure.

We proposed a novel neural network based on stochastic/statistical theory, and applied it to the development of learning control technology in robots and medical welfare equipment and for the classification of medical data.

In this study, we proposed a markerless infant movement assessment system to evaluate general movements (GMs) in infants using video images.

We developed an electromyogram (EMG) prosthetic hand that has five independently driven fingers. The hand includes a flexion drive and a force-magnification drive, allowing rapid finger motion and a firm grasp.

In this study, a novel stochastic model for surface electromyogram (EMG) signals was proposed. In the model, the variance of EMG signals is handled as a random variable, thereby allowing the estimation of signal-dependent noise that cannot be represented in previous EMG models.