Mingshi Wang

Research on Features of Retinal Images Associated with Hypertension and Diabetes

In order to discuss the feasibility of detecting hypertension and diabetes using certain features of retinal images, this paper compared features of twenty-six retinal images from three groups, normotensive, hypertension and diabetes. Those features including exudates, bifurcation angle, artery-to-veins diameter ratio, mean artery and vein diameters, shape and size of optic disc, and vessel tortuosity. The results show that in retinal images of hypertension and diabetes, exudates appeared, and the area increased with years of illness. The mean arterial diameter tended to decrease while the mean vein diameter dilated. Compared to the normotensive group, the artery-to-vein diameter ratio decreased for both the hypertensive and diabetic group. The shape and size changes of the optic disc were obvious in the diabetic group. The hypertensive group showed tortuous blood vessels. No tortuous blood vessels were observed in the diabetic group. The study demonstrates that, expect the bifurcation angles, it is unfeasible to use retinal image features in the detection of hypertension and diabetes, but they do have some meaning in clinic diagnosis of hypertension and diabetes since some characteristics of retinal images of those patients become pathologically changed

White Matter Fiber Tracking Method by Vector Interpolation with Diffusion Tensor Imaging Data in Human Brain

The directional diffusion information obtained by diffusion tensor magnetic resonance imaging (DT-MRI) can be used to noninvasively visualize white matter fibers in human brain. And it is the unique way to reconstruct those fibers in vivo. However, there are some obstacles for the reconstruction process, e.g., partial volume effect caused by low spatial resolution negatively affects the DT-MRI results, and continuous fibers are indicated by the discrete signal on the DT-MRI. A fiber tracking algorithm is proposed in this paper. Major eigenvectors and fractional anisotropy (FA) values of the correlative voxels are used for fiber tracking. As a result, this method shows stronger potential to depict the distribution of white matter fibers in human brain

Identification and Classification for finger movement based on EEG

Identification and classification technology plays an important part in study of the BCI system. There are many algorithms to classify the event of different task related. Here, finger movement was used as the basic and typical tasks to be identified in the BCI experiments. The ideas of BP and ERD were introduced and discussed. The CSSD (common spatial subspace decomposition) algorithm was used for classifying single-trial EEG during the preparation of left-right finger movements after the two kinds of phenomena were expounded in detail in this paper. Experiment and simulating results show that the averaged classification accuracy can be up to the 75.6%

Study of Feature Classification Methods in BCI Based on Neural Networks

Feature classification is one of the important aspects in brain-computer interfaces (BCI) system. It has been known that a higher precision can be achieved if use neutral networks in a proper way for feature classification. In this paper, three feature identification ways were introduced and discussed. In the experiment of left-right hand classification, the arithmetic of the small mean square difference is proposed and studied, so as to get a good converging in the task classification. The design method of input and output layer for the BP neural network was discussed. Experiment results show that it is a feasible processing algorithm to classify the different events

Influence of Time-Varying Magnetic Field on the Release of Neurotransmitters in Raphe Nuclei of Rats

A specially-designed time-varying magnetic field was developed to treat insomnia. Clinical results showed that this method could shorten the time to go to sleep and prolong the sleep duration. However, the mechanism of this method is still not well understood. In this study, the effect of magnetic stimulation on the release of serotonin (5-HT), noradrenaline (NE), dopamine (DA) in raphe nuclei of rats, which are known to play an important role in the sleep-wake regulation, was investigated. It was shown that there was a significant difference in the release of serotonin between control group and experimental group (p < 0.01). The release of serotonin of the experimental group increased significantly. No obvious release changes of NE and DA are found (p > 0.05). The results indicates that one possible mechanism of inducing sleep using specially designed magnetic field is to change the release of sleep-related neurotransmitters

Study of the Electromagnetic Field of Transcranial Magnetic Stimulation Based on the Real Head Model

This paper presents the application of finite element method in analysis of the electromagnetic field within real head model exposed to a time varying magnetic field produced by the circuit-driving loop. To analyze the electromagnetic field under Transcranial Magnetic Stimulation precisely and display the stimulating effect of the loops in different geometrical forms, a 3-D real head model reconstructed from the CT Images is used to simulate the head. Based on the practical circuit of the magnetic stimulating instrument, the cerebral electromagnetic field produced by the loops of 4 different forms is analyzed with their electromagnetic performance evaluated. The work can provide the academic and practical basis for further study upon Transcranial Magnetic Stimulation and developing more complicated loop or loop array.

Study on effect and mechanism of magnetic fields simulating EEG rhythm upon memory

Magnetic fields simulating EEG rhythm were used to stimulate Wistar rats to explore the effect of magnetic field on retrieval (recall) ability and its mechanism. The results indicated that most of the weak magnetic fields (>10 minutes) simulating the EEG rhythm of human brain impaired the retrieval of long-term memory significantly (P<0.05), but weak magnetic field with special rhythm may even have the capability of facilitating memory performance. And the effects of TMS on memory can last for at least several hours (5h) after TMS. Compared with control group, the release of NE, DA and 5-HT in hippocampus of stimulated group increased (P<0.05); While the release of ACh decreased (P<0.05). Through electronic microscope, morphological changes of nerve synapses in hippocampus of rats were observed after weak magnetic stimulation. The percentage of alpha and beta rhythm in EEG power spectra changed in cats after induced by weak magnetic fields simulating the EEG rhythm.

Event-related brain potentials during visual sentence reading and picture recognition memory tasks

It is uncertain if different brain areas in response to pre-semantic picture processing are functionally homogeneous. Using event-related potentials (ERPs), we aimed to explore the neural activities in different brain regions in relation to processing of sentence memory and picture identification. Healthy subjects were chosen to discriminate visual stimulus pairs, and the ERPs were recorded from the scalp. Two kinds of stimuli were provided for each subject in the present study. One was Chinese sentence reading, referred as task 1. Another one was watching a line-drawing picture to judge if the picture matched the meaning of the sentence before. When the line-drawing picture received by the subject was inconsistent with the meaning of the sentence before, it was called as task 2, otherwise, if incongruous, it was called as task 3. Our findings implicate that stimuli of sentence memory and picture identification may exert neural activities on different working memory areas in the brain of human.

Preliminary study on mechanism of the induction of sleep by time-varying magnetic field

Before and after magnetic stimulation, the discharge frequency of raphe nuclei in rabbits was recorded. The results show that there is obvious change of the discharge frequency, t<0.01. This presents one of the possible mechanisms of the induction of sleep by time-varying magnetic field.

An Automatic Optimum Data selection Method For EEG-based Brain-computer Interface

An electroencephalogram (EEG) based brain-computer interface (BCI) is aimed at developing a system that can support communication possibilities for patients with severe neuromuscular disabilities through EEG pattern recognition and classification. Previously many parametric modeling techniques for EEG analysis have been developed and improved upon. For this work we analyzed five parameters on seven subjects to study their influence on brain computer interface (BCI) classification. Our study shows that these parameters greatly influence classification accuracy with subject dependent parameters. This suggests that the parameter selection process should be analyzed further when building models.