Takamasa Shimada

Detection of characteristic waves of sleep EEG by neural network analysis

In psychiatry, the sleep stage is one of the most important forms of evidence for diagnosing mental disease. However, doctors require much labor and skill for diagnosis, so a quantitative and objective method is required for more accurate diagnosis since it depends on the doctors experience. For this reason, an automatic diagnosis system must be developed. In this paper, the authors propose a new type of neural network (NN) model referred to as a sleep electroencephalogram (EEG) recognition neural network (SRNN) which enables one to detect several kinds of important characteristic waves in sleep EEG which are necessary for diagnosing sleep stages. Experimental results indicate that the proposed NN model was much more capable than other conventional methods for detecting characteristic waves.

Sensor Network Management for Healthcare Applications

Telemedicine can be implemented using a sensor network that carries users’ biometric information collected by tiny intelligent sensors attached to the human body. We propose a communication system that can be used for telemedicine to improve the Quality of Life of individuals, which uses a sensor network and a sensor database with mutual authentication to ensure security. Finally, this paper describes the priority control mechanism and experiments using it and clarifies the real-time performance of the proposed communication sensor network system.

Estimation of temporary change of activation areas by moving an analysis time window in fMRI measurement

In this paper, we propose a method to acquire temporal changes of activations by moving an analysis time window. An advantage of this method is that it can acquire rough changes of activated areas even with the data having low time resolution. We ascertained that activations from our method do not contradict previous reports on the oddball paradigm, thus showing its effectiveness. Eight normal subjects participated in the study, which consisted of a random series of 30 target and 70 nontarget stimuli. We investigated the activated area in three kinds of analysis time sections, from stimulus onset to 5 s after the stimulus (time section A), from 2 to 7 s after (B) and from 4 to 9 s after (C). In time section A, representative activated areas were regions including the left and supplementary motor areas (SMA), and cerebellum. In B, regions including the left motor area and SMA, right parahippocampal gyrus (Broadmann Area (BA) 30), right limbic lobe and cerebellum were activated. In C, bilaterally postcentral gyrus (BA 3,40), right anterior cingulate (ACC, BA 32), left middle frontal gyrus (BA 9) and right parahippocampal gyrus were activated. Most activations were consistent with previous studies.

Human brain mapping of judgment system of movement of visualized material

Human visual system has been frequently studied. Anatomical and physiological investigations indicate functional streams within the extrastriate visual cortex of the macaque monkey. It shows that V5 area participates in motion perception. However, the region participating motion judgment was not very investigated. In this paper, we tried to reveal the region of motion judgment. We used functional magnetic resonance imaging (fMRI) to explore. The short videos of moving ball (lasting 1s) and stationary ball (lasting 1s) were randomly presented to subjects repeatedly. Each paradigm consisted of two main conditions (control and activation condition). During control, subjects continued tapping regardless of stimuli. During activation condition, subjects tap only when the video of moving ball was presented (motion judgment). When subjects judge motion, activity was revealed in frontal cortex (Brodmann 9).

Analysis of P300 evoked by multisensory stimulus consisting of visual and auditory stimuli

In this study, we investigated P300 amplitude and latency when multisensory stimuli are presented, and compared the results with those in a single sensory stimulus. Here, visual and auditory stimulus were used as sensory stimulus, and these stimuli were presented to the subject simultaneously in the case of multisensory stimuli. In the experiment, three kinds of stimuli, visual, auditory, and multisensory stimuli were prepared, and either one of visual and auditory stimulus or both were set as target stimulus. The subject was instructed to push a button when target stimulus occurred. As a results, P300 amplitude of grand averaged waveform in visual evoked potential (VEP) changed a lot depending on the kind of target stimulus. The other hand, P300 amplitudes in auditory evoked potential (AEP) were almost same regardless of the kind. Furthermore, the amplitude in VEP and AEP was decreasing in the following order: target stimulus is consistent with the corresponding sensory stimulus; both visual and auditory stimuli are the target; the target is inconsistent with a sensory stimulus; and no target is set. The amplitude in multisensory evoked potential (MEP) was decreasing in the following order: both visual and auditory stimuli are the target; the target is auditory stimulus; the target is visual stimulus; no target is set. A significant difference was recognized in P300 amplitudes between visual target and no target. Regarding P300 latency of grand averaged waveform in VEP and AEP, the latencies with target stimulus were longer than that without target. However, in the case of MEP, there was rare difference in the existence of target stimulus. As for response time until button pushing, in MEP, the time when target is both of visual and auditory stimuli tended to be shorter than that when target is single sensory stimulus though significant difference could not be recognized.

Evaluation of brain aging by EEG analysis of photic driving response

In this paper, we examined the possibility to estimate brain aging by EEG photic driving response. Normal subjects are classified into 3 age groups by age, 20s, 21 – 59, and over 60. We obtained the Z-score from EEG at rest and during photic stimulus (PS) by using the method for evaluating intraindividual EEG we have already proposed. Here, we showed the averaged Z-score of 3 age groups for 6 Hz PS as a representative example. As a results, in 20s, significant difference (|Z|>1.96, p<0.05) was recognized at fundamental and higher harmonic frequencies of PS frequency except second harmonic. With increase of age, we could see the decrease of Z-score at fundamental frequency and higher harmonics. In the group over 60, no significant difference at any frequencies was recognized.

Brain Activation in Response to Disgustful Face Images with Different Backgrounds

Previous studies have demonstrated that the stimulus of a fearful and disgustful face images lead to activation of neural responses in the medial temporal lobe. In particular, it was reported that seeing disgustful face image activated the insula area of brain. In these studies, no background images were used with facial stimuli. However, normal daytoday images always have a background. Moreover, background images are considered important in art forms (painting, photography, and movies, etc.) for eliciting effective expressions. We assessed the effect of background images on brain activation by using functional magnetic resonance imaging (fMRI). During fMRI scanning, face images with background images were presented repeatedly to 8 healthy righthanded males. Facial stimuli comprised 5 photographs of a disgustful face selected from the Paul Ekman’s database (Ekman and Friesen 1976). The background images comprised 2 photographs-one is worms and the other is a flower garden. It is thought that disgustful face images coincide with worms background image on the point of impression. After scanning, the subjects rated the impression created by the images on the Plutchik scale. Significant effects of the image of the disgustful face against the worms background minus that against the flower garden were assessed using a ttest and displayed as statistical parametric maps (SPMs) using SPM2 software. The results demonstrated activation of the right insula, and the image of the disgustful face against the worms background created a more disgustful impression than that against the flower garden. Therefore, the image of the face and the background together create the overall impression. The difference in the activation of the insula is possibly induced by the creation of this overall impression. This demonstrates the importance of background images in forming an impression of face images.

FMRI Analysis of the Effect of Background Image Combined with Face Image

In previous studies, activations in neural responses in the medial temporal lobe have been found with the stimuli of fear face. Especially the amygdala is thought to be part of a neural system responsive to potential threat. But previous these studies have used no background images for facial stimuli. In the present study, we estimated the effect of background image to the brain activation by using fMRI. During fMRI scanning, face image with background image were presented repeatedly. The background images consisted of two photographs, lightning and flower garden. After the fMRI scanning session, significant effects (P<0.001) with the contrast (fearful face with lightning) minus (fearful face with flower garden) were assessed using a t-test with SPM2 software. As a result, the activation of amygdala and hippo-campus were found. This fact shows the importance of background image on giving impression to person with face images.

The temporal change of activations in the visual target oddball stimulus

In recent years, many researchers have been studying activated regions in the brain using fMRI to investigate the source of brain activity for various sensory stimuli. However, previous reports have not necessarily been consistent. We examined the activation for a visual target task in the oddball paradigm as a first step to investigating sources associated with the functions of decision and recognition. In our analysis, we examined the activated area in the brain in three kinds of analysis time sections, from stimulus onset to 5 seconds after the stimulus (time section A), from 2 to 7 seconds (time section B) and from 4 to 9 seconds (time section C). In the temporal change, in the time section A, many activated areas could be seen. On the whole, the flow of activation from the parietal area activated in time section A to the prefrontal area activated in the time section B and C could be seen. It is considered from our results that various areas are involved in the visual target response.

Fast EEG spike detection via eigenvalue analysis and clustering of spatial amplitude distribution

OBJECTIVE In the current study, we tested a proposed method for fast spike detection in electroencephalography (EEG). APPROACH We performed eigenvalue analysis in two-dimensional space spanned by gradients calculated from two neighboring samples to detect high-amplitude negative peaks. We extracted the spike candidates by imposing restrictions on parameters regarding spike shape and eigenvalues reflecting detection characteristics of individual medical doctors. We subsequently performed clustering, classifying detected peaks by considering the amplitude distribution at 19 scalp electrodes. Clusters with a small number of candidates were excluded. We then defined a score for eliminating spike candidates for which the pattern of detected electrodes differed from the overall pattern in a cluster. Spikes were detected by setting the score threshold. MAIN RESULTS Based on visual inspection by a psychiatrist experienced in EEG, we evaluated the proposed method using two statistical measures of precision and recall with respect to detection performance. We found that precision and recall exhibited a trade-off relationship. The average recall value was 0.708 in eight subjects with the score threshold that maximized the F-measure, with 58.6  ±  36.2 spikes per subject. Under this condition, the average precision was 0.390, corresponding to a false positive rate 2.09 times higher than the true positive rate. Analysis of the required processing time revealed that, using a general-purpose computer, our method could be used to perform spike detection in 12.1% of the recording time. The process of narrowing down spike candidates based on shape occupied most of the processing time. SIGNIFICANCE Although the average recall value was comparable with that of other studies, the proposed method significantly shortened the processing time.