HTML5 Webook

17/64

NICT REPORT 15money (500 yen in this case), while our subject decides whether to accept or re-ject the proposal. If the proposal is accept-ed, the money is distributed according to the proposal, but if it is rejected, both par-ties receive nothing. Earlier research has shown that most people would reject a proposal if their allocated portion was 20% or less. In this research, subjects made 56 decisions for differing proposers and pro-posals. Subjects also repeated the Beck Depression Inventory II one year later to study their depressive tendencies. We at-tempted to predict depressive tendencies, current and one year later, based on pat-terns of activity in the amygdala and hip-pocampus arising from differences when a proposal was offered, using Bayesian sparse regression with a kernel function (which is a machine learning method). We observed a significant positive correlation between the predicted and measured val-ues (Fig.2). This positive correlation be-tween predicted and measured values shows the potential for prediction. This re-sult also hints at the importance to a per-son’s mental state, of comparisons with other people (disparity), and shows that information processing in the amygdale and hippocampus is an underlying factor.Explaining brain function and next generation ICT research is-suesIn this research, we focused on a phe-nomenon called illusory jitter, in which ob-jects that are not actually moving appear to shake with a frequency of approximately 10 cycles per second. We first confirmed that differences among subjects in the rhythm of the perceived jitter reflect differ-ences in the rhythm of their alpha brain waves. In other words, people with faster alpha wave rhythm tended to perceive faster jitter, while people with slower alpha wave rhythm perceived slower jitter. We also developed a technology able to artifi-cially change the rhythm of alpha waves by applying weak, non-harming electrical stimulation to the back of the subject’s head (transcranial stimulation). When this technology was used to change the rhythm of alpha waves, the illusory jitter perceived by the subjects also changed similarly (Fig.3). These results verify that the rhythm of alpha waves contribute to perception of illusory jitter, and suggest that the timing for synthesis of information about shape and movement, which are processed in different locations in the brain, is determined by alpha waves. Brain function measurement tech-nologiesGenerally, the signals originating from brain activity that can be observed using fMRI are extremely weak, so the accuracy of the data is usually increased statistically by taking repeated measurements with an imaging technique called single-shot EPI, that has excellent time resolution. Howev-er, 7T-fMRI can produce stronger signals than earlier technologies, so we changed from the conventional approach of priori-tizing time resolution, and used multi-shot EPI, which increases the accuracy of the signals collected themselves. We also in-creased the efficiency of measurements by rearranging how the tasks were given, and by capturing structured images, which have excellent histological contrast, we built an fMRI test system able to make de-tailed comparisons of structure and func-tion. As a result, we were able to collect fMRI data with spatial resolution of 0.6 mm (Fig.4).Fig.3 : When alpha waves are changed by electrical stimulation, the rhythm of illusory jitter also changesFig.4 : Ultra-high-resolution fMRI under visual stimulation with 0.6 mm voxels (capturing the brain activity localized in the visual cortical structure)Fig.2 : (Left) Part of the brain related to economic ineq-uity. (Right) Relationship between changes in depres-sion tendencies one year later and values predicted from brain activity patterns.Research and Development