Dissertation > Medicine, health > Basic Medical > Medical science in general > Biomedical Engineering > General issues > Biological information, biological control

The Research of Finger -Movement-Related MEG Signal Processing

Author LiWeiNa
Tutor HouWenSheng
School Chongqing University
Course Biomedical Electronics and Information Technology
Keywords Magnetoencephalography (MEG) Artifact removal Time-frequency power spectrum expressed (TFRs) Event-related synchronization / desynchronization ( ERD / ERS ) Function source extraction (FSS)
CLC R318.04
Type Master's thesis
Year 2009
Downloads 89
Quotes 2
Download Dissertation

The hand is mankind through the long-term evolution of the formation of the most dexterous movement organs, from the hand seen in the projected area of ??the motor cortex, the brain of the manual control mechanism is very complex. With millisecond time resolution and millimeter spatial resolution magnetoencephalography (MEG), as a non-invasive detection of brain electromagnetic physiological signal brain function detection technology, undoubtedly to study the cortex of good manual control mechanism means. Explore the control mechanisms of the brain opponent not only in itself is a very significant thing, the results can also be applied to the clinical diagnosis of sports functional diseases, rehabilitation, treatment and rehabilitation evaluation, while the mechanism prosthetic control, Dexterous Manipulator control important application value. Based on the premise of magnetoencephalography to explore the brain mechanisms of manual control is able to effectively extract the characteristics of the tasks related magnetoencephalography signal. Therefore this article to explore the mechanisms of the brain control finger movement oriented, following research carried out for the purpose of a preliminary understanding MEG signal processing methods and the characteristics of the finger movement related cortical neural activity related to system finger movement. MEG signal processing method, the paper first discusses the MEG signal preprocessing focus MEG signal artifact removal method. By ICA model component screening module artifact method to improve the traditional ICA to visit two aspects from the the component statistical characteristics and spectral characteristics of each component to filter out the artifact component, the ultimate realization of MEG artifact of automatically removed. Followed by the MEG signal time-frequency power spectrum analysis. In this paper, the frequency power spectrum signal in-depth study of the theory (TFRs) for low-band signal variable window length, single-window time-frequency power spectrum estimation and variable window length, multi-window for the high-band signal frequency power spectrum estimation, and in particular the finger buttons related to the MEG signal the mu rhythm beta rhythm and low gamma rhythms when frequency power spectrum analysis. Finally, when the earlier MEG signal frequency power spectrum characteristics as functional constraints ICA's goal function to extract finger keystrokes cortical neural activity source function of source extraction method (FSS). Through the the evoked MEG signal when the case of normal human index finger buttons frequency power spectrum analysis, and get the following results: (1) finger buttons related MEG signal energy is mainly concentrated in the low frequency band below 35Hz, 40Hz to 70Hz band cortical neural activity, but lower energy bands weak; (2) the low frequency rhythm wave (mu rhythms and beta rhythms) in the preparation and execution of finger keystrokes during exhibit power suppression (ERD) and beta rhythm in the action execution after instantaneous power sudden and substantial increase (ERS), duration about one second, and the mu rhythm suppression status is still in power after the action execution; keystrokes before the preparation and execution, (3) high frequency rhythms (especially low gamma band rhythm) no significant activities in the action-executive and executive showed significant power enhancement; (4) the rhythm wave activity location: mu rhythm bilateral central and parietal regions. channel activities have corresponding cortical sensorimotor area, strong-side activity in the ipsilateral; beta rhythmic activity is mainly concentrated in the channel of contralateral central and parietal regions compared to the active area in front of the button and the active area of ??the button slightly in front, corresponding to the primary motor area; low gamma rhythmic activity in double contralateral activity stronger than the ipsilateral side of the central area channel. Band neural activity ERD and ERS physiological significance of the discussion can be speculated: cortical sensorimotor area mu rhythm ERD is the sign of the sensorimotor cortex of the brain activated; beta rhythm ERD and the ERS can be characterized by the movement of the finger from the preparation, execution to restore the entire process after exercise; low gamma band ERS may be associated with the integration of a variety of information processing. Extracted by calculating this article finger keystrokes evoked MEG signal source, with CTF brain magnetic system SAM software to extract the correlation between the magnetic source, verify both significantly correlated, thus proving the extract function source. Function source activities are mainly concentrated in the trailing edge of the channel of contralateral central area, roughly corresponding to the front side of the central sulcus central gyrus cortical motor area. Function source time-domain waveform analysis, key experimental design has not been like somatosensory evoked magnetic field as the characteristic wave, but the FSS Act to obtain the time domain information from a large number of samples explore the time evolution of the brain control finger movement time-frequency characteristics of the mechanism and analysis functions source basis.

Related Dissertations
More Dissertations