This article summarizes the special talk entitled “Beginning of words emerges from caregiver-infant interactions” at 55th Scientific Meeting of the Japanese Society of Stomatognathic Function. First, cognitive developmental robotics (hereafter, CDR) which the speakers have been advocating for more than 15 years was introduced. The key ideas of CDR are “physical embodiment,” and “sociality” which can be seamlessly connected through “development.” Symbol emergence also shares these ideas, especially, in case of speech perception and vocalization since the process from random cooing and babbling to vowel utterances of infants can be regarded as a process of symbol emergence through the interaction with their mothers. Next, computational approaches are shown, especially focusing on mother-infant speech interaction. Not only the learning algorithms on infant (robot) side, but also how mothers interact with their infants, which could be a sort of entrainment, is discussed. Simulations based on a sensori-motor magnet bias and an auto-mirroring one show how these biases affect the process of vowel acquisition of infant. Finally, future issues are shown.

View full abstract

In elderly dominated society, a robotics technology is expected to support the elder, the disable, the patient and the healthy people. In this article, the wearable arm robot for the patients with essential tremor, rollover support orthosis for the patient with cancer bone metastasis and the haptic navigation interface for the patient with the visually impaired are introduced to discuss the possibility of robotics technology based on the collaboration of medicine and engineering.

View full abstract

Although daytime clenching is believed to be one of the oral parafunctions leading to dental problems, a treatment strategy has not yet been devised. Electromyogram (EMG) biofeedback (BF) training was performed to ascertain its effect on regulating daytime clenching behavior. However, the long-term effects remain unknown. Therefore, the purpose of this study was to determine the learning effect concerning one month after EMG-BF training.
 Twelve subjects (7 males, 5 females: mean age; 31.5 ± 6.3 years) who had mild masticatory muscle pain with daytime clenching behavior were divided into the BF group (n=6) and control (CO) group (n=6) in a random fashion. Subjects were fitted with a hearing-aid-shaped EMG biofeedback apparatus, which was used to record EMG data under natural conditions from the temporal muscle, continuously for five hours on four consecutive days and 1 month after (1 Mo). EMG data on Days 1, 4 and 1 Mo were recorded without biofeedback as the pre-test, post-test, and follow-up test, respectively. On Days 2 and 3, subjects in the BF group were alerted to their clenching behaviors via an alert sound from the EMG biofeedback apparatus. No alert sound was given to the CO group throughout the recording sessions.
 No significant difference was observed in the number of clenching events for five hours between the BF group (7.3 ± 5.1) and CO group (7.5 ± 2.1) on Day 1; however, significant decreases were found in the BF group between Day 1 (7.3 ± 5.1) and Day 4 (2.3 ± 1.5; p < 0.05), and Day 1 and 1 Mo (2.0 ± 1.1; p < 0.05; Wilcoxon with Bonferroni's inequality test).
 Daytime clenching was reduced in the long term with the help of an EMG biofeedback system under natural circumstances. Daytime EMG-BF training decreased clenching events, and the effect may continue for one month after training.

View full abstract