基于腦電和肌電相干性的輔助中風(fēng)病人主動康復(fù)方法研究
發(fā)布時間：2019-03-04 15:07
【摘要】：中風(fēng)是一種高死亡率和高致殘率的腦血管疾病。大部分的中風(fēng)幸存患者會喪失很多功能,諸如運(yùn)動功能、語言功能、記憶功能、視覺功能等�，F(xiàn)代康復(fù)醫(yī)學(xué)研究認(rèn)為康復(fù)訓(xùn)練對中風(fēng)患者的功能康復(fù)具有重大的促進(jìn)作用。因此研究有效的康復(fù)訓(xùn)練方法對中風(fēng)患者的恢復(fù)具有重大的實(shí)際意義。中風(fēng)患者的康復(fù)可以分為被動康復(fù)和主動康復(fù)兩大類。在被動康復(fù)中,中風(fēng)患者的主動意識并沒有參與到康復(fù)訓(xùn)練中。然而在主動康復(fù)中,中風(fēng)患者自主意識主動參與到康復(fù)訓(xùn)練中。研究表明主動康復(fù)的效果比被動康復(fù)效果要理想。另外中風(fēng)患者手部功能在日程生活中占有很大的比重,如何恢復(fù)手部功能具有重要意義。 為了中風(fēng)患者手部功能的康復(fù),本文以建立有效的主動康復(fù)方法為研究目標(biāo)。在分析了關(guān)鍵技術(shù)和難點(diǎn)以后,本文從兩個方面出發(fā)分別提出有效的解決方法,1)建立了一種基于肌電模型調(diào)制的電刺激系統(tǒng)用于輔助手部功能康復(fù)方法；2)在此基礎(chǔ)上本文基于腦電肌電相干性(Cortico-muscular coherence (CMC))分析,提取患者的自主意識信號,建立了一套有效的主動康復(fù)方法,用于手部功能的康復(fù)訓(xùn)練。 本文根據(jù)健康被試抓握不同直徑不同重量物體時發(fā)放的肌電信號建立一個肌電模型,克服了抓握不同物體時肌電信號難以表達(dá)的問題；通過該肌電模型來調(diào)制電刺激系統(tǒng)形成一個相應(yīng)的電刺激模型,使該電刺激模型刺激強(qiáng)度和正常人肌電發(fā)放模型相一致,這與傳統(tǒng)恒壓或者恒流電刺激相比,通過肌電模型調(diào)制的電刺激模型可以有效延緩肌肉疲勞,延長使用時間。由于中風(fēng)患者肌肉存在肌痙攣、肌萎縮、肌無力等癥狀,導(dǎo)致單純利用肌電信號不能有效的提取出自主意識信號以區(qū)分不同的動作。另一方面腦電信號對手部精細(xì)動作的區(qū)分率不高,有效的提取出被試的自主意識信號也存在一定的問題。考慮到上述的不足,本文對基于現(xiàn)有的肌電信號(Electromyography (EMG))、腦電信號(electroencephalography (EEG))以及腦電肌電相干性信號(CMC)提取自主意識信號進(jìn)行了重點(diǎn)研究。本文對EMG信號、EEG信號和CMC信號分類準(zhǔn)確率進(jìn)行了對比,發(fā)現(xiàn)在健康被試進(jìn)行指伸vs.拇內(nèi)收,指伸vs.指屈,指伸v s.靜息手部不同動作時,單純用EEG腦電信號的分類準(zhǔn)確率分別為71.7_+12.32%,71.2_+12.80%,81.39_+11.52%；而采用CMC分類準(zhǔn)確率則分別為78.,6_+4.29%,81.0_+7.34%和78.025_+9.39%。在中風(fēng)患者進(jìn)行指伸v s.指屈,指伸vs.靜息手部不同動作時,單純用EMG肌電信號的分類準(zhǔn)確率為分別為68.79_+0.41%,97.42_+0.27%；單純用EEG腦電信號的分類準(zhǔn)確率分別為81-0_+0.53%,85-9_+0.23%；而利用CMC腦電肌電相干性的分類準(zhǔn)確率則分別為85.44_+1.06%,91.87+132%。健康被試的CMC與EEC信號相比,指伸vs.拇內(nèi)收,指伸vs.指屈提高了10%左右。中風(fēng)患者的CMC與EEG相比,指伸vs.指屈,指伸vs.靜息分別提高了3.54%和5.98%。CMC與EMG相比,指伸vs.指屈提高了16.65%。以上研究結(jié)果證明了無論是健康被試還是中風(fēng)患者與單純使用EEG信號或單純使用EMC信號相比,CMC能更有效的提取自主意識信號。 根據(jù)以上的研究基礎(chǔ),本文設(shè)計(jì)了一個用于中風(fēng)患者手部功能康復(fù)的系統(tǒng)。該系統(tǒng)以利用CMC信號用于準(zhǔn)確提取中風(fēng)患者主動意識信號,該主動意識信號有效的用于控制外部的多通道電刺激系統(tǒng)設(shè)備輔助中風(fēng)患者的手部功能康復(fù)。 縱觀本文的研究成果,主要創(chuàng)新點(diǎn)在于(1)以健康被試抓握不同直徑、不同重量物體時的肌電信號為依據(jù),建立了一個肌電模型,對手部不同動作進(jìn)行了定量的表達(dá),通過肌電模型調(diào)制電刺激強(qiáng)度建立一個電刺激模型,可以用于中風(fēng)患者的電刺激(electrical stimulation,Es)治療,相比恒定強(qiáng)度的電刺激方法,該電刺激可以延緩肌肉疲勞,增加使用時間；(2)研究表明,基于腦電肌電相干性信號(CMC信號)與BEG信號,EMG信號相比可以提高分類準(zhǔn)確率,能有效的提取自主意識信號。 本研究為中風(fēng)患者手部的主動康復(fù)提供客觀的依據(jù),有利于系統(tǒng)的進(jìn)一步研制。
[Abstract]:Stroke is a high-mortality and high-disability-rate cerebrovascular disease. Most of the stroke survivors lose a lot of their functions, such as exercise, language, memory, vision, and so on. The study of modern rehabilitation medicine considers that the rehabilitation training has a great effect on the function rehabilitation of patients with stroke. Therefore, the effective rehabilitation training method is of great practical significance to the recovery of stroke patients. The rehabilitation of patients with stroke can be divided into two categories: passive rehabilitation and active rehabilitation. In passive rehabilitation, the active consciousness of stroke patients is not involved in the rehabilitation training. However, in the active rehabilitation, the self-consciousness of stroke patients is actively involved in the rehabilitation training. The results show that the effect of active rehabilitation is more than that of passive rehabilitation. In addition, that hand function of the stroke patient has a great specific gravity in the schedule life, and how to restore the hand function is of great significance. In order to recover the hand function of patients with stroke, this paper aims to establish an effective method of active rehabilitation for the purpose of study. On the basis of the analysis of the key technology and the difficult point, this paper puts forward an effective solution to solve the problem from two aspects, and 1) a kind of electric stimulation system based on the myoelectric model modulation is set up to assist the rehabilitation of the hand function. Methods:2) Based on the analysis of the corico-muscular coherence (CMC), the self-consciousness signal of the patient was extracted, a set of effective methods of active rehabilitation and the rehabilitation training for hand function were established. In this paper, a myoelectric model is set up according to the electromyographic signals that are distributed in different weight objects with different diameters, and the problem that the myoelectric signal is difficult to express when the different objects are grasped is overcome. The electric stimulation system is modulated by the myoelectric model to form a corresponding electric spike. The stimulation model of the electric stimulation model is consistent with the normal human myoelectric distribution model, and compared with the traditional constant-voltage or constant-current electric stimulation, the electric stimulation model which is modulated by the myoelectric model can effectively delay the muscle fatigue and prolong the time The symptoms of muscle spasm, amyotrophy, myasthenia gravis and other symptoms in the muscle of the stroke patient can not be effectively extracted by using the myoelectric signal to distinguish the difference of the self-consciousness signal. and on the other hand, the differentiation rate of the fine movement of the brain electrical signal opponent part is not high, and the effective extraction of the self-consciousness signal to be tested also exists In the light of the above-mentioned deficiencies, the present paper is based on Electromyography (EMG), Electroencephalography (EEG) and Electromyoelectric Coherence Signal (CMC) to extract the self-awareness signal. In this paper, the classification accuracy of EMG signal, EEG signal and CMC signal is compared, and it is found that the classification accuracy of EEG signal is 71.7 __ + 12.32%, 71.2 __ + 12.80%, 81.39 __ + 11, respectively. and the classification accuracy of the CMC is respectively 81.0 _ + 7.34% and 78.025 _ + 9 for 78,6 _ + 4.29%, 81.0 _ + 7.34% and 78.025 _ + 9. The classification accuracy of EMG signal was 68.79 _ + 0.41%, 97.42 _ + 0.27%, respectively. The classification accuracy of EEG was 81-0 _ + 0.53% and 85-9 _ + 0, respectively. The accuracy of the classification of the CMCs was 85.44, 1.06%, 91.87 + 1, respectively. 32%. The CMC of the healthy subjects compared with the EEC signal, the finger flexion and extension vs. the finger flexion increased by 10 compared to the EEC signal. The CMC of stroke patients increased by 3.54% and 5.98%, respectively, compared with the EEG. 65%. The results of the above study demonstrate that CMC can be more effective in extracting a self-mind, both in healthy subjects and in stroke patients compared to simply using an EEG signal or simply using an EMC signal Understanding the signal. Based on the above research basis, this paper designs a hand function for stroke patients. The system is used for accurately extracting a stroke patient active consciousness signal by using a CMC signal, and the active consciousness signal is effective for controlling an external multi-channel electric stimulation system device to assist a stroke patient's hand The main innovation point of this paper is to (1) establish a myoelectric model based on the electromyographic signals of different diameters and different weight objects in healthy subjects. the quantitative expression is performed, the electrical stimulation intensity is modulated by the myoelectric model to establish an electric stimulation model, and the electric stimulation model can be used for the treatment of the electrical stimulation (Es) of a stroke patient, and compared with the electric stimulation method with the constant intensity, the electric stimulation can delay the muscle fatigue and increase the use time; (2) The study shows that the classification accuracy can be improved compared with that of the BEG signal and the EMG signal based on the EEG signal (CMC signal). This study provides an objective basis for the active rehabilitation of the hand of a stroke patient.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級別】：博士
【學(xué)位授予年份】：2014
【分類號】：R743.3
[Abstract]:Stroke is a high-mortality and high-disability-rate cerebrovascular disease. Most of the stroke survivors lose a lot of their functions, such as exercise, language, memory, vision, and so on. The study of modern rehabilitation medicine considers that the rehabilitation training has a great effect on the function rehabilitation of patients with stroke. Therefore, the effective rehabilitation training method is of great practical significance to the recovery of stroke patients. The rehabilitation of patients with stroke can be divided into two categories: passive rehabilitation and active rehabilitation. In passive rehabilitation, the active consciousness of stroke patients is not involved in the rehabilitation training. However, in the active rehabilitation, the self-consciousness of stroke patients is actively involved in the rehabilitation training. The results show that the effect of active rehabilitation is more than that of passive rehabilitation. In addition, that hand function of the stroke patient has a great specific gravity in the schedule life, and how to restore the hand function is of great significance. In order to recover the hand function of patients with stroke, this paper aims to establish an effective method of active rehabilitation for the purpose of study. On the basis of the analysis of the key technology and the difficult point, this paper puts forward an effective solution to solve the problem from two aspects, and 1) a kind of electric stimulation system based on the myoelectric model modulation is set up to assist the rehabilitation of the hand function. Methods:2) Based on the analysis of the corico-muscular coherence (CMC), the self-consciousness signal of the patient was extracted, a set of effective methods of active rehabilitation and the rehabilitation training for hand function were established. In this paper, a myoelectric model is set up according to the electromyographic signals that are distributed in different weight objects with different diameters, and the problem that the myoelectric signal is difficult to express when the different objects are grasped is overcome. The electric stimulation system is modulated by the myoelectric model to form a corresponding electric spike. The stimulation model of the electric stimulation model is consistent with the normal human myoelectric distribution model, and compared with the traditional constant-voltage or constant-current electric stimulation, the electric stimulation model which is modulated by the myoelectric model can effectively delay the muscle fatigue and prolong the time The symptoms of muscle spasm, amyotrophy, myasthenia gravis and other symptoms in the muscle of the stroke patient can not be effectively extracted by using the myoelectric signal to distinguish the difference of the self-consciousness signal. and on the other hand, the differentiation rate of the fine movement of the brain electrical signal opponent part is not high, and the effective extraction of the self-consciousness signal to be tested also exists In the light of the above-mentioned deficiencies, the present paper is based on Electromyography (EMG), Electroencephalography (EEG) and Electromyoelectric Coherence Signal (CMC) to extract the self-awareness signal. In this paper, the classification accuracy of EMG signal, EEG signal and CMC signal is compared, and it is found that the classification accuracy of EEG signal is 71.7 __ + 12.32%, 71.2 __ + 12.80%, 81.39 __ + 11, respectively. and the classification accuracy of the CMC is respectively 81.0 _ + 7.34% and 78.025 _ + 9 for 78,6 _ + 4.29%, 81.0 _ + 7.34% and 78.025 _ + 9. The classification accuracy of EMG signal was 68.79 _ + 0.41%, 97.42 _ + 0.27%, respectively. The classification accuracy of EEG was 81-0 _ + 0.53% and 85-9 _ + 0, respectively. The accuracy of the classification of the CMCs was 85.44, 1.06%, 91.87 + 1, respectively. 32%. The CMC of the healthy subjects compared with the EEC signal, the finger flexion and extension vs. the finger flexion increased by 10 compared to the EEC signal. The CMC of stroke patients increased by 3.54% and 5.98%, respectively, compared with the EEG. 65%. The results of the above study demonstrate that CMC can be more effective in extracting a self-mind, both in healthy subjects and in stroke patients compared to simply using an EEG signal or simply using an EMC signal Understanding the signal. Based on the above research basis, this paper designs a hand function for stroke patients. The system is used for accurately extracting a stroke patient active consciousness signal by using a CMC signal, and the active consciousness signal is effective for controlling an external multi-channel electric stimulation system device to assist a stroke patient's hand The main innovation point of this paper is to (1) establish a myoelectric model based on the electromyographic signals of different diameters and different weight objects in healthy subjects. the quantitative expression is performed, the electrical stimulation intensity is modulated by the myoelectric model to establish an electric stimulation model, and the electric stimulation model can be used for the treatment of the electrical stimulation (Es) of a stroke patient, and compared with the electric stimulation method with the constant intensity, the electric stimulation can delay the muscle fatigue and increase the use time; (2) The study shows that the classification accuracy can be improved compared with that of the BEG signal and the EMG signal based on the EEG signal (CMC signal). This study provides an objective basis for the active rehabilitation of the hand of a stroke patient.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級別】：博士
【學(xué)位授予年份】：2014
【分類號】：R743.3
【共引文獻(xiàn)】
中國期刊全文數(shù)據(jù)庫 前10條
1 劉榮梅;;早期康復(fù)訓(xùn)練對腦梗死患者偏癱恢復(fù)的影響[J];當(dāng)代醫(yī)學(xué);2010年34期
2 常娜;劉森;;針灸結(jié)合康復(fù)訓(xùn)練治療肩手綜合征的療效分析[J];第四軍醫(yī)大學(xué)學(xué)報(bào);2005年24期
3 尤春景;肌肉疼痛綜合征的診斷和治療(續(xù)前)[J];國外醫(yī)學(xué)(物理醫(yī)學(xué)與康復(fù)學(xué)分冊);1997年01期
4 謝戰(zhàn)杰;王子鴻;鄔弋;;168例傷害死亡病例流行病學(xué)特征分析[J];中國公共衛(wèi)生管理;2014年03期
5 朱現(xiàn)民;侯靜s，
本文編號：2434379
本文鏈接：http://www.sikaile.net/yixuelunwen/shenjingyixue/2434379.html
最近更新
教材專著
