• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.481-481
• /
• 2000

We investigated the structure-activity relationships on use of multi-layer neural networks. The relationships are techniques required in developments of medicines. Since many kinds of observations might be adopted on the techniques, we discussed some points between the observations and the properties of multi-layer neural networks. In the structure-activity relationships, an important property is not that standard deviations are nearly equal to zero for observed physiological activity, but prediction ability for unknown medicines. Since we adopted non-linear approximation, the function to represent the activity can be defined by observations; therefore, we believe that the standard deviations have not significance. The function was examined by "leave-one-out" method, which was originally introduced for the multi-regression analysis. In the linear approximation, the examination is significance, however, we believe that the method is inappropriate in case of nonlinear fitting as neural networks; therefore, we derived a new index fer the relationships from the differential of information propagation in the neural network. By using the index, we discussed physiological activity of an anti-cancer medicine, Mitomycine derivatives. The neuro-computing suggests that there is no direction to extend the anti-cancer activity of Mitomycine, which is close to the trend of anticancer developing.

• Journal of Information Processing Systems
• /
• v.14 no.1
• /
• pp.150-161
• /
• 2018

We propose a novel deep neural network model for detecting human activities in untrimmed videos. The process of human activity detection in a video involves two steps: a step to extract features that are effective in recognizing human activities in a long untrimmed video, followed by a step to detect human activities from those extracted features. To extract the rich features from video segments that could express unique patterns for each activity, we employ two different convolutional neural network models, C3D and I-ResNet. For detecting human activities from the sequence of extracted feature vectors, we use BLSTM, a bi-directional recurrent neural network model. By conducting experiments with ActivityNet 200, a large-scale benchmark dataset, we show the high performance of the proposed DeepAct model.

• Korean Journal of Acupuncture
• /
• v.38 no.3
• /
• pp.162-174
• /
• 2021

Objectives : We aimed to identify the antidepressant effect of liver tonification acupuncture treatment (ACU (LT); KI10, LR8, LU8, LR4) and four gate acupuncture treatment (ACU (FG); LI4, LR3) and its brain neural activity in the normal and chronic restraint stress (CRS)-induced mouse model. Methods : Firstly, normal mice were given ACU (LT) or ACU (FG) and the c-Fos expressions in each brain region were analyzed to examine brain neural activity. Secondly, CRS was administered to mice for 4 weeks, then ACU (LT) or ACU (FG) was performed for 2 weeks. The depression-like behavior was evaluated using open field test (OFT) before and after acupuncture treatment. Then, the c-Fos expressions in each brain region were analyzed to examine brain neural activity. Results : In normal mice, ACU (FG) regulated brain neural activities in the hypothalamus, hippocampus, and periaqueductal gray. ACU (LT) changed more brain regions in the prefrontal cortex, insular cortex, striatum, and hippocampus, including those altered by ACU (FG). In CRS-induced model, ACU (LT) alleviated depression-like behavior more than ACU (FG). Also, brain neural activities in the motor cortex area 2 (M2), agranular ventral part and piriform of insular cortex (AIV and Pir), and cornu ammonis (CA) 1 and CA3 of hippocampus were changed by ACU (LT), and those of AIV and CA3 were also changed by ACU (FG). As in normal mice, ACU (LT) resulted in changes in more brain regions, including those altered by ACU (FG) in CRS model. M2, Pir, and CA1 were only changed by ACU (LT) in depression model, suggesting that these brain regions reflect the specific effect of ACU (LT). Conclusions : ACU (LT) relieved depression-like behavior more than ACU (FG), and this acupuncture effect was associated with modulation of brain neural activities in the motor cortex, insular cortex, and hippocampus.

• Journal of International Academy of Physical Therapy Research
• /
• v.9 no.2
• /
• pp.1447-1454
• /
• 2018

The objective of this study was to investigate the dynamic neural mobilization program on the changes in muscle activity and nerve conduction velocity (NVC) in stroke patients. The participants were sampled and randomly divided into experimental group I (n=12) who underwent arm neural mobilization and experimental group II (n=13) who underwent arm dynamic neural mobilization. As the pretest, peripheral NVC of the radial, median, and ulnar nerves were measured using the Viking Quest; the biceps brachii, brachioradialis, flexor carpi radialis, and extensor carpi radialis activities were measured with sEMG. Each intervention program consisted of 10 trials per set and three sets per session. The intervention programs were performed once daily for four weeks (four days/week). Posttest measurements were taken equally as the pretest measurements. Significant differences in peripheral NVC in all sections of the radial and median nerves and wristbelow elbow and below elbow-above elbow areas of the ulnar nerve, as well as in muscle activity of all muscles except the biceps brachii. These findings indicate that dynamic neural mobilization was effective in increasing peripheral NVC and altering the muscle activity.

• Proceedings of the Korean Society for the Gifted Conference
• /
• 2003.05a
• /
• pp.137-138
• /
• 2003

Understanding how and why people differ is a fundamental, if distant, goal of research efforts to bridge psychological and biological levels of analysis. General fluid intelligence (gF) is a major dimension of individual differences and refers to reasoning and novel problemsolving ability. A conceptual integration of evidence from cognitive (behavioral) and anatomical studies suggeststhat gF should covary with both task performance and neural activity in specific brain systems when specific cognitive demands are present, with the neural activity mediating the relation between gF and performance. Direct investigation of this possibility will be a critical step toward a mechanistic model of human intelligence. In turn, a mechanistic model might suggest ways to enhance gF through targeted behavioral or neurobiological intervent ions, We formed two different groups as subjects based on their scholarly attainments. Each group consists of 20 volunteers(aged 16-17 years, right-handed males) from the National Gifted School and a local high school respectively. To test whether individual differences in general intelligence are mediated at a neural level, we first assessed intellectual characteristics in 40 subjects using standard intelligence tests (Raven's Advanced Progressive Matrices, Wechsler Adult Intelligence Scale, Torrance Tests of Creative Thinking) administered outside of the MR scanner. We then used functional magnetic resonance imaging (fMRl) to measure task-related brain activity as participants performed three different kinds of computerized reasoning tasks that were intended to activate the relevant neural systems. To examine the difference of neural activity according to discrepancy in general intelligence, we compared the brain activity of both extreme groups (each, n=10) of the participants based on the standard intelligence test scores. In contrast to the common expectation, there was no significant difference of brain region involved in high-g tasks between both groups. Random effect analysis exhibited that lateral prefrontal, anterior cingulate and parietal cortex are associated with gF. Despite very different task contents in the three high-g-low-g contrasts, recruitment of multiple regions is markedly similar in each case, However, on the task with high 9F correlations, the Prodigy group, (intelligence rank: >99%) showed higher task-related neural activity in several brain regions. These results suggest that the relationship between gF and brain activity should be stronger under high-g conditions than low-g conditions.

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.2
• /
• pp.90-96
• /
• 2023

Securing large amounts of training data in deep learning neural networks, including convolutional neural networks, is of importance for avoiding overfitting phenomenon or for the excellent performance. However, securing labeled training data in deep learning neural networks is very limited in reality. To overcome this, several augmentation methods have been proposed in the literature to generate an additional large amount of training data through transformation or manipulation of the already acquired traing data. However, unlike training data such as images and texts, it is barely to find an augmentation method in the literature that additionally generates bio-signal training data for convolutional neural network based human activity recognition. Thus, this study proposes a simple but effective augmentation method of bio-signal training data for convolutional neural network based human activity recognition. The usefulness of the proposed augmentation method is validated by showing that human activity is recognized with high accuracy by convolutional neural network trained with its augmented bio-signal training data.

• The Journal of Korean Medicine
• /
• v.24 no.3
• /
• pp.51-64
• /
• 2003

Objective : This study investigated the alteration of neural activity and effect of Yanggyuksanhwa-tang (Lianggesanhuo-tang) on cerebral ischemia of rats. Methods : Considering age-related impact on cerebral ischemia, aged rats (18 months old) were used for this study. Ischemic damage was induced by the transient occlusion of bilateral common carotid arteries (BCAO) with hypotension. Yanggyuksanhwa-tang (Lianggesanhuo-tang) was administered twice a day orally. Then alterations of neural activities in the brain of aged BCAO rats were measured by the [$^{14}C$]2-deoxyglucose autoradiography method. Results : The BCAO in aged rats led to significant decrease of neural activity in the whole brain. Treatment with Yanggyuksanhwa-tang (Lianggesanhuo-tang) significantly attenuated the decrease of neural activity in the whole brain following BCAO ischemia. Treatment significantly attenuated the decrease of neural activity in the CA1, CA2, CA3, dentate gyrus of the hippocampus, activated barrel, barrel cortex, somatosensory cortex, cingulate cortex, caudate putamen, and medial septal nucleus following BCAO in aged rats. Treatment with Yanggyuksanhwa-tang (Lianggesanhuo-tang) also significantly attenuated the decrease of neural activity in the anteroventral thalamic nucleus, ventral anterior thalamic nucleus, arcuate nucleus, posterior hypothalamic area, medial mammillary nucleus, lateral periaqueductal gray, dorsal raphe nucleus, interpeduncular nucleus, median raphe nucleus, and medial pontine nucleus. Conclusion : It can be suggested that Yanggyuksanhwa-tang (Lianggesanhuo-tang) has a neuroprotecuve effect on cerebral ischemia through the control of glucose metabolic rate and cerebral blood flow.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1996.10a
• /
• pp.39-42
• /
• 1996

Modeling techniques such as linear regression have been used to predict hurricane activity many months in advance of the start of the hurricane season with some success. In this paper, we construct feedforward neural networks to model Atlantic basin hurricane activity and compare the predictions of our neural network models to the predictions produced by statistical models found in the weather forecasting literature. We find that our neural network models produce reasonably accurate predictions that, for the most part, compare favorably to the predictions of statistical models.

• Proceedings of the Korean Society for Cognitive Science Conference
• /
• 2010.05a
• /
• pp.55-59
• /
• 2010

To examine iterocular interactions in normal subjects, we recorded EEG activity from channel O1 and O2 on 14 healthy subjects while checkerboard pattern reversals were presented to each eye with different interstimulus intervals (ISIs) ranging from 0 to 218 ms. When pattern reversals were presented asynchronously to each eye, P-1 activity evoked by each reversal was significantly suppressed compared to the activation evoked by synchronous reversals. Furthermore, when there was time delay between pattern reversals to each eye, theta (4-10 Hz) band power was also significantly decreased, whereas beta (10-30 Hz) band power was relatively preserved. Our results suggest that neural activity evoked by sustained but not yet reversed checkerboard from one eye might inhibit upcoming neural response evoked by reversed checkerboard from the other eye. Decreased P-1 activity might reflect such inhibitory interaction between two eyes. Also, decreased theta and preserved beta band power might reflect engagement of different neural circuit for binocular / monocular vision.

• Development and Reproduction
• /
• v.19 no.2
• /
• pp.61-67
• /
• 2015

The immediate early gene c-fos has long been known as a molecular marker of neural activity. The neuron's activity is transformed into intracellular calcium influx through NMDA receptors and L-type voltage sensitive calcium channels. For the transcription of c-fos, neural activity should be strong enough to activate mitogen-activated protein kinase (MAPK) signaling pathway which shows low calcium sensitivity. Upon translation, the auto-inhibition by Fos protein regulates basal Fos expression. The pattern of external stimuli and the valence of the stimulus to the animal change Fos signal, thus the signal reflects learning and memory aspects. Understanding the features of multiple components regulating Fos signaling is necessary for the optimal generation and interpretation of Fos signal.