We measured brain activity to investigate how accurately music needs to be imagined to obtain similar relaxation effects to listening to actual music. Prefrontal activity was determined from changes in oxy- and deoxy-Hemoglobin (Hb), measured by a 10-channel wearable optical topography instrument. Participants were allowed to assess their imagination accuracy for music intuitively, using the visual analog scale, defined in this manuscript as imagining-accuracy (IA). We observed relaxation during preferred music known to the participant and preferred music unknown to the participant. We found the IA for music needs to be approximately 80% to obtain the same relaxation level as listening to music. This suggests that participants know music is an important factor for relaxation and that relaxation can increase by repeating unknown music. We also discovered that some participants showed decreased brain activity by imagining music in their own way.

The role of mucosal immunity in the defense against pathogens is well established. However, there does not seem to be much research on the relationship between salivary secretory immunoglobulin A (SIgA), activities of daily living (ADL), and cognitive function, particularly among geriatric residents of long-term health care facilities. For this, the cognitive function and SIgA concentrations of 49 residents of such a long-term health care facility in Japan were evaluated across 3mobility groups, namely, the walking (n = 11), wheelchair (n = 19), and bedridden groups (n = 19).

Bedridden residents had lower mini-mental state than examination scores indicative of moderate cognitive impairment and significantly higher SIgA concentrations than the wheelchair and walking groups. We concluded that enhanced dedicated care of the mobile or partially mobile geriatric residents may be an instrument in infection prophylaxis.

Screening of fundus images is an important phase for an automatic analysis of fundus image, where they are classified into “good” or “bad” fundus images. The convolutional neural network (CNN) is a typical classification method, but with use of CNN, we cannot decide which feature contributes to the resulting classification. The purpose of this paper is to propose a screening method, keeping accuracy of CNN, to classify the fundus images into “good” or “bad” images using the features that can explain the classification results. In the proposed method, the gray level co-occurrence matrix (GLCM) together with other color space statistics of RGB and HSV are employed. Accuracy rate of the proposed method was almost the same as that of CNN, but the extracted features by the proposed method could explain the classification results. The features employed in the proposed method could give the reasons of the “bad” fundus images.

In recent years, sound quality control using active noise control has been used as a noise reduction method, targeting low frequency noise. Specifically, an algorithm for amplifying and reducing the engine-specific order components is proposed. In addition, the auditory impressions of engine sound controlled by the proposed method were investigated using psychoacoustic measurements. The scale values of preference for each stimulus were obtained by Scheffe’s paired comparison tests. When the reduction level increased, the preference was decreased or increased from the reference sound. Further, when the amplification-level increased, the preference was decreased from the reference sound. These results indicated that the control corresponding to the individual is important for improvements in auditory impressions. To solve these problems, the proposed method was developed to adjust to individual preferences. The individual preferences of sound were connected to each driver’s driving pattern using deep learning. Thus, the proposed system enables the automatic generation of individual sound preferences.

Considering the increasing number of patients with disabilities, especially of upper extremity, we decided to develop hand rehabilitation virtual reality system for post-stroke patients using leap motion controller. However, as the device has not met the required accuracy of hand position tracking, we created a real-time low-computational cost correction algorithm lessening the problem. This paper outlines the challenges of hand position tracking accuracy identifying the parts contributing to the overall error, proposes a method of correction suitable for each of the parts separately and brings the parts together into an algorithm maximizing the hand position correction.

The algorithm was tested on twelve healthy participants in two series of four tasks. Results showed our algorithm improved the hand position accuracy considerably in all the cases.