微电子神经桥接实现蟾蜍腿部运动功能重建

doi:10.3969/j.issn.1673-5374.2013.06.008

摘要/Abstract

摘要：

实验以神经信号探测电极阵列、功能电激励用电极阵列和微电子电路（包括放大器、信号处理器和功能电激励驱动电路）构建微电子神经桥。对1只脊蟾蜍左腿进行外部机械刺激或硫酸化学刺激，通过2只脊蟾蜍间的微电子神经桥在另一只脊蟾蜍左腿上实现神经信号和腿部动作的功能重建。利用深踪示波器发现受控脊蟾蜍的肌电信号是由控制脊蟾蜍的神经信号引起的，二者之间有一定的延迟。结果证明微电子神经桥可以用来重建受损神经的功能。

关键词: 微电子神经桥, 神经功能重建, 肌电信号, 相干函数, 神经损伤, 脊髓反射弧, 脊蟾蜍, 神经电生理, 再生, 神经再生

Abstract:

The present study used a microelectronic neural bridge comprised of electrode arrays for neural signal detection, functional electrical stimulation, and a microelectronic circuit including signal amplifying, processing, and functional electrical stimulation to bridge two separate nerves, and to restore the lost function of one nerve. The left leg of one spinal toad was subjected to external mechanical stimulation and functional electrical stimulation driving. The function of the left leg of one spinal toad was regenerated to the corresponding leg of another spinal toad using a microelectronic neural bridge. Oscilloscope tracings showed that the electromyographic signals from controlled spinal toads were generated by neural signals that controlled the spinal toad, and there was a delay between signals. This study demonstrates that microelectronic neural bridging can be used to restore neural function between different injured nerves.

Key words: neural regeneration, basic research, microelectronic neural bridge, electromyographic signal, coherence function, nerve injury, spinal reflex arc, spinal toad, grants-supported paper, photographs-containing paper, neuroregeneration

. 微电子神经桥接实现蟾蜍腿部运动功能重建[J]. 中国神经再生研究(英文版), 2013, 8(6): 546-553.

Xiaoyan Shen, Zhigong Wang, Xiaoying Lv, Zonghao Huang. Microelectronic neural bridging of toad nerves to restore leg function[J]. Neural Regeneration Research, 2013, 8(6): 546-553.

参考文献

[1] Giszter SF. Spinal cord injury: present and future therapeutic devices and prostheses. Neurotherapeutics. 2008;5(1):147-162.

[2] Tanriverdi T, Al-Jehani H, Poulin N, et al. Functional results of electrical cortical stimulation of the lower sensory strip. J Clin Neurosci. 2009;16(9):1188-1194.

[3] Vinit S, Kastner A. Descending bulbospinal pathways and recovery of respiratory motor function following spinal cord injury. Respir Physiol Neurobiol. 2009;169(2): 115-122.

[4] Bamford JA, Todd KG, Mushahwar VK. The effects of intraspinal microstimulation on spinal cord tissue in the rat. Biomaterials. 2010;31(21):5552-5563.

[5] Bamford JA, Mushahwar VK. Intraspinal microstimulation for the recovery of function following spinal cord injury. Prog Brain Res. 2011;194:227-239.

[6] Johnson MD, Franklin RK, Gibson MD, et al. Implantable microelectrode arrays for simultaneous electrophysiological and neurochemical recordings. J Neurosci Methods. 2008;174(1):62-70.

[7] Stieglitz T, Rubehn B, Henle C, et al. Brain-computer interfaces: an overview of the hardware to record neural signals from the cortex. Prog Brain Res. 2009;175: 297-315.

[8] Do AH, Wang PT, King CE, et al. Brain-computer interface controlled functional electrical stimulation system for ankle movement. J Neuroeng Rehabil. 2011;8:49.

[9] Gad P, Woodbridge J, Lavrov I, et al. Forelimb EMG-based trigger to control an electronic spinal bridge to enable hindlimb stepping after a complete spinal cord lesion in rats. J Neuroeng Rehabil. 2012;9(1):38.

[10] Fairchild MD, Kim SJ, Iarkov A, et al. Repetetive hindlimb movement using intermittent adaptive neuromuscular electrical stimulation in an incomplete spinal cord injury rodent model. Exp Neurol. 2010;223(2):623-633.

[11] Bittar CK, Cliquet A Jr. Rehabilitation programme using neuromuscular electrical stimulation in spinal cord: epidemiological aspects. Acta Med Port. 2011;24(4): 527-532.

[12] Lai CH, Chang WH, Chan WP, et al. Effects of functional electrical stimulation cycling exercise on bone mineral density loss in the early stages of spinal cord injury. J Rehabil Med. 2010;42(2):150-154.

[13] Velez FG, Isobe J, Zealear D, et al. Toward an implantable functional electrical stimulation device to correct strabismus. J AAPOS. 2009;13(3):229-235.e1.

[14] Courtine G, Gerasimenko Y, van den Brand R, et al. Transformation of nonfunctional spinal circuits into functional states after the loss of brain input. Nat Neurosci. 2009;12(10):1333-1342.

[15] Cooke RM, Parker D. Locomotor recovery after spinal cord lesions in the lamprey is associated with functional and ultrastructural changes below lesion sites. J Neurotrauma. 2009;26(4):597-612.

[16] Hakansson NA, Hull ML. Muscle stimulation waveform timing patterns for upper and lower leg muscle groups to increase muscular endurance in functional electrical stimulation pedaling using a forward dynamic model. IEEE Trans Biomed Eng. 2009;56(9):2263-2270.

[17] Gwilliam JC, Horch K. A charge-balanced pulse generator for nerve stimulation applications. J Neurosci Methods. 2008;168(1):146-150.

[18] Bütikofer R, Lawrence PD. Electrocutaneous nerve stimulation-II: stimulus waveform selection. IEEE Trans Biomed Eng. 1979;26(2):69-75.

[19] Wang ZG, Lv XY, Li WY, et al. Study of microelectronics for detecting and stimulating of central neural signals. In: IEEE proceedings of international conference on neural interface and control. Wuhan, 2005.

[20] Wang ZG, Gu XS, Lv XY. Neural channel bridge aided by a micro-electronic system. China, CN200510135541.6. 2006-08-02.

[21] Wang ZG, Gu XS, Lv XY. Neural channel bridge aided by a micro-electronic system. USA, 8000806. 2011-08-16.

[22] Wang ZG, Gu XS, Lv XY, et al, Microelectronics-embedded channel bridging and signal regeneration of injured spinal cords. Prog Nat Sci. 2009; 19(10):1261-1269.

[23] Ichihara K, Venkatasubramanian G, Abbas JJ, et al. Neuromuscular electrical stimulation of the hindlimb muscles for movement therapy in a rodent model. J Neurosci Methods. 2009;176(2):213-224.

[24] Judge SE. General anaesthetic action in the invertebrate central nervous system. Gen Pharmacol. 1980;11(4): 337-341.

[25] Judge SE, Norman J. The action of general anaesthetics on acetylcholine-induced inhibition in the central nervous system of Helix. Br J Pharmacol. 1982;75(2):353-357.

[26] Shen XY, Wang ZG, Lu XY, et al. Neural signal sensing, transmission and functional regeneration on different toads' bodies. Conf Proc IEEE Eng Med Biol Soc. 2009; 2009:5953-5956.

[27] Shen XY, Wang ZG, Lu XY, et al. Neural function rebuilding on different bodies using microelctronic neural bridge technique. Dongnan Daxue Xuebao. 2010;26(4): 523-527.

[28] Akin T, Najafi K, Smoke RH, et al. A micromachined silicon sieve electrode for nerve regeneration applications. IEEE Trans Biomed Eng. 1994;41(4):305-313.

[29] Richardson RR Jr, Miller JA, Reichert WM. Polyimides as biomaterials: preliminary biocompatibility testing. Biomaterials. 1993;14(8):627-635.

[30] Haggerty HS, Lusted HS. Histological reaction to polyimide films in the cochlea. Acta Otolaryngol. 1989; 107(1-2):13-22.

[31] Brimble MJ, Musabayane CT. Blood-circulated sciatic nerve-gastrocnemius muscle preparation in the spinal toad. Physiologist. 1984;27(1):47-49.

[32] De Michele G, Sello S, Carboncini MC, et al. Cross-correlation time-frequency analysis for multiple EMG signals in Parkinson's disease: a wavelet approach. Med Eng Phys. 2003;25(5):361-369.

[33] Subasi A, Yilmaz M, Ozcalik HR. Classification of EMG signals using wavelet neural network. J Neurosci Methods. 2006;156(1-2):360-367.

[34] Strambi SK, Rossi B, De Michele G, et al. Effect of medication in Parkinson's disease: a wavelet analysis of EMG signals. Med Eng Phys. 2004;26(4):279-290.

[35] Yianni J, Wang SY, Liu X, et al. A dominant bursting electromyograph pattern in dystonic conditions predicts an early response to pallidal stimulation. J Clin Neurosci. 2006;13(7):738-746.

[36] Mallat SG. A Wavelet Tour of Signal Processing. Maryland Height: Academic Press. 2009.

[37] Samar VJ. Wavelet analysis of neuroelectric waveforms Brain Lang. 1999;66(1):1-6.

[38] Grosse P, Cassidy MJ, Brown P. EEG-EMG, MEG-EMG and EMG-EMG frequency analysis: physiological principles and clinical applications. Clin Neurophysiol. 2002;113(10):1523-1531.

[39] Zhan Y, Halliday D, Jiang P, et al. Detecting time-dependent coherence between non-stationary electrophysiological signals--a combined statistical and time-frequency approach. J Neurosci Methods. 2006; 156(1-2):322-332.

[40] Issartel J, Marin L, Gaillot P, et al. A practical guide to time-frequency analysis in the study of human motor behavior: the contribution of wavelet transform. J Mot Behav. 2006;38(2):139-159.

引言

材料方法

Design

A self-controlled study.

Time and setting

The experiments were performed at Nantong University and Southeast University, China between July 2009 and December 2011.

Materials

Animals

Spinal toads, weighing 45–55 g, were purchased from Shengming Research Animal Farms, Nanjing, China. Toads were kept in indoor round aquariums, with 20 cm depth of well water or aerated tap water over night. Water temperature varied between 12–25°C. A small quantity of pebbles and stones was used for barricades and caves. Three 20 watt fluorescent bulbs were used as the light source, and a water submersible pump was used to add oxygen, and a heater used to heat. All expermental procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals, published by the National Institute of Health.

Instrumentation

Composition of the microelectronic neural bridge: In principle, a neural bridge is used to provide a multi-channel link across injured nerve fibres to bypass interrupted neural signals^[26-27]. The block diagram of our microelectronic neural bridge is shown in Figure 3. It consists of an electrode array, a head stage, a filtering network, a preamplifier, a main amplifier, an functional electrical stimulation driver, and another electrode array. The first electrode array is used to detect the neural signal from the proximal stump of injured nerve fibres. The head stage, the filtering network, and the preamplifier preliminarily amplify the neural signal, but limits interference. After further amplification by the main amplifier, the neural signal is sent to the functional electrical stimulation driver, which supplies a high voltage or a strong current to the second electrode array so that the desired neural signal can be regenerated in the distal nerve stump.

It is challenging to detect a neural signal from a neuron or nerve fibre because the signal is relatively weak and there is considerable noise and interference. Noise for the first stage should be as low as possible. Because the impedance of the signal source, i.e. the organism, is high and uncertain, the equivalent current noise is the main consideration. Therefore, a unit-gain buffer is used as the head stage. The filtering network is made up of a low-pass filter that provides attenuation to high- frequency noise and interference before the preamplifier. To reduce the common-mode interference in the lower frequency band, a differential input system with a 128-dB common mode reject ratio was designed. For the purpose of noise reduction, operational amplifiers with an ultra-low noise figure were adopted (Figure 4). The system’s 3-dB bandwidth is from 10 Hz to 2 kHz. The output current of the functional electrical stimulation driver is adjustable from 0 to 7.8 mA.

Electrodes: Two types of electrodes, hooked electrodes (Quanshui Experimental Devices Co., Nanjing, China) and self-made needle electrode arrays, were used. The hooked electrodes were made of two parallel metal wires pressed into the surface layer of a hook-formed plastic plate. To meet the electrochemical requirements and to get “smooth” geometry for the prevention of mechanical trauma and irritation caused by sharp edges, polyimide was used as the substrate and insulation material. Polyimide has been shown to be a non-toxic material suitable for neural implantation^[28-30]. The self-made needle electrodes arrays consisted of acupuncture needles covered with Parylene C (Cookson Electronics Specialty Coating Branch, Shanghai, China) except at the tips. The needles with exposed tips were thin and suitable for selective activation of motor nerves.

Methods

Establishment of the spinal toad model

Before experiments, two spinal toads were prepared^[31]. The toads were stunned by a blow to the head. A fine scalpel blade was inserted into the spinal canal at the back of the skull severing the brain from the spinal cord. A blunt needle was then inserted through the hole made by the scalpel blade and pushed forward into the skull via the foramen magnum to destroy the brain. A longitudinal incision was made through the skin of the dorsal side of the upper leg, and the sciatic nerve exposed by pulling apart the muscle bundles. The sciatic nerve was freed from surrounding tissue, with care taken to avoid damage to the femoral artery and vein that run alongside the nerve. The sciatic nerve was exposed between the biceps femora and semimembranosus muscles. A small wad of absorbent cotton soaked in Ringer solution was placed over the nerve to prevent drying. The spinal toads recovered from spinal shock within 10 minutes.

The experimental set-up is depicted in Figure 5. The hooked electrodes were in contact with the sciatic nerve of the source spinal toad, which was stimulated by external stimulation. The cable connector was connected to the input of the microelectronic neural bridge. Simultaneously, the neural signal detected from the sciatic nerve was monitored and recorded using an oscilloscope (Agilent 64200 or Agilent DSO6014A, CA, USA). Another hooked electrode connected with the output of the microelectronic neural bridge contacted the sciatic nerve of the controlled spinal toad. The self-made acupuncture needle electrodes were fixed to the optimal sites for monitoring of electromyographic signals: at the left legs of both spinal toads and under the bellies of the gastrocnemius muscles.

Data processing

The signals acquired by the electrodes were monitored and digitally recorded by the oscilloscope (Agilent 64200 or Agilent DSO6014A, USA). The recorded data were analysed offline using Matlab (MathWorks, MA, USA). Before the analysis of functional coupling between muscular activity and simultaneously recorded oscillatory neural activities of the motor system, denoising and pattern recognition were carried out using wavelet transformation. Wavelet transformation provides a method of decomposing signals both in scale and time. During the last decade, wavelet transformation has been used in different areas of engineering, biology, and medicine^[32-34]. Because of its superiority in analysing transient and non-stationary signals, it has become a powerful tool and extension of Fourier analysis. Most biomedical signals are not stationary and have highly complex time-frequency characteristics. The wavelet method can provide good time resolution at high frequency and good scale resolution at low frequency^[35]. The wavelet function used was DB3 wavelets^[36]. After wavelet transformation, pattern recognition was completed according to the amplitude and time course of the spikes. Finally, raster graphics were obtained, and neural spikes extracted.

The analysis of functional coupling between muscular activities, while simultaneously recording oscillatory neural activity at different levels of the motor system, has provided significant insights into the neural control of movement^[37-39]. One widely used method of estimating the similarity, synchronism, or functional coupling between two oscillatory signals is the coherence function ρ_xy, defined by^[40]

where x(n) and y(n) are discrete signals; m is the number of shift points and delay time; m > 0 means that sequence y(n) shifts left, while m < 0 means that sequence y(n) shifts right.

The coherence function is a normalized correlation function, and measures the relative linearity and delay time between two signals. The value of coherence has no relationship with the magnitude of oscillatory signals, and is convenient for comparing the degree of relativity. It can not only express phase coherence, but can also express phase-shift (or delayed) coherence. The coherence value indicates the strength of the coupling in the time domain. It is mathematically bounded between 1 and +1, where 1 indicates a perfect linear relationship and 0 indicates no linear relationship.

The coherent function was calculated not only between the sciatic neural signals of source spinal toad and controlled spinal toad, but also between electromyographic and sciatic neural signals. The phase differences provided an estimation of the temporal delays between those signals.

讨论

文章快阅

延伸阅读

实验背景：

重建由于损伤或者疾病引起的运动或感觉功能是临床神经科学面临的最严峻挑战之一。神经科学、生物医学和电子科学交叉领域的发展，使得重建受损神经的功能在技术上实现创新成为可能。其中一种方案是提高受损神经轴突的再生能力或者用其他细胞来替换，另外一种方案是采用模式发生器，通过功能电激励使损伤点以下的神经通路恢复功能。这种功能电激励常常被用来修复瘫痪病人的运动功能。目前的功能电激励系统都是采用外部的电脉冲信号激发损伤点以下的神经网络，从而产生期望的运动。已经有研究表明在脊髓损伤患者的硬脊膜外施加刺激可以产生固定模式的类运动式的动作。刺激信号的参数，如电激励的频率对不同动物会有所不同。尽管目前在神经功能再生方面已经取得了一些进展，获得了一些新的认识，但还有很多难题尚未解决。

以上提到的功能电激励有个共同的特性，那就是其激励脉冲信号都是人工设计的信号。与之不同的是，我们这里不需要外加信号，而是采用称之为微电子神经桥的集成电路桥接于受损神经两端，从而重建丧失的神经功能。

微电子神经桥的概念是在2005年首次提出的。我们已经设计了几代混合集成电路和单片集成的微电子神经桥，开展了一系列的动物实验。通过50只大鼠的实验，验证了近端坐骨神经信号或脊髓神经信号可以在远端坐骨神经或脊髓上得到再生。以家兔为动物模型，在截断的脊髓神经两端实现了双向桥接，自发的神经信号和诱发的神经信号都实现了再生，观察到了相应的信号波形，从方法上验证了微电子神经桥的可行性。

但在大鼠和兔子的实验中，存在2个方面的局限性：(1)电极植入时需要对动物进行麻醉。虽然麻醉能使大鼠或家兔保持安静，但会抑制神经的活性。(2)再生时是采用简单脉冲电激励的诱发方式，与此相对应的只是观察到简单的动作如抽动或抖动。

鉴于第1个局限性，课题组引入了脊蟾蜍这种动物模型。脊蟾蜍是去脑或损毁了大脑但脊髓反射弧保持完好的蟾蜍。引入脊蟾蜍有2个优点：一是不需要麻醉，因此不会抑制神经信号的产生和传播，也不会抑制神经功能电激励产生相应的神经信号。二是作为两栖动物，蟾蜍整个躯体可以在不受大脑控制的情况下较长时间保持生物活性。这2点都非常有利于神经再生实验的顺利进行。

鉴于第2个局限性，微电子神经桥的信号源取自没有任何伪迹的与腿动作相对应的坐骨神经信号。我们知道，动作是受神经信号控制的。在没有任何外部激励的情况下，只有当微电子神经桥将信源脊蟾蜍坐骨神经信号成功地提取出来并传输到受控的脊蟾蜍，受控脊蟾蜍才能做出与信源脊蟾蜍相对应的腿部动作。文章在机械刺激和化学刺激下，在2只分离的脊蟾蜍之间通过微电子神经桥实现了腿功能重建，从而证明，利用微电子神经桥可以重建协调的腿动作。