视网膜仿真模型及其感知效能分析
|
摘要
生物学家Rosen说过：“从进化的观点来说,生理系统是人类解决复杂问题的最好的百科全书”。而人的眼睛,它叹为观止的完善功能,复杂精细的组织结构,精密协调的控制机制更是“极其完美和复杂的”(达尔文),可以说是百科全书中最为绚丽的章节之一。我们的工作正是受启发于眼睛中接收和处理信号最为重要的组织—视网膜。
很早之前,人们就已认识到了视网膜的复杂,但具体复杂到什么程度,视网膜可分为哪些层,每层有哪些种细胞,每种细胞的数量和分布如何,则是随着近几年来解剖学,电生理学,细胞形态学等学科的发展而逐渐清晰起来的。基于这些数据,我们建立了一个高度逼真的视网膜计算模型,此模型不仅模拟了视网膜的多种细胞类型和分布,还拟合了视网膜复杂的层次连接和结构,既忠于生理上的结构和处理过程又在整体上展现了视网膜的多个信息处理通路。在模型基础上可以设计很多实验,比如视网膜物体检测概率实验,视网膜物体表征效率实验,视网膜在多条件限定下(如准确性,实时性,能耗,计算复杂度等)的均衡性分析实验等等。
有了这个高度逼真的视网膜计算模型,可以为当前很多棘手的问题提供解决的方向。如人工视网膜芯片的设计。人工视网膜因为尺寸有限,所以硬件复杂度不能太高；因为发热问题,所以能耗不能太大；因为要满足实时的反应要求,所以信号处理需要快速且准确。这些约束条件相互冲突,很难平衡。但这些工作正是视网膜平时所做的事情。研究视网膜在多条件下的平衡性质,对人工视网膜的设计具有巨大的指导意义。此模型也可用于实时的场景表征和图像处理。视网膜将信号分为颜色,亮度,运动等不同的通路形成内部表征再进行进一步处理,这在提高表征的效率的同时大幅降低了计算的复杂度,为实时图像的表征和处理提供了一个新的方向。同时,由于模型高度逼真于真实的视网膜,也可以进行生理实验,验证生理上的假设和推测,为生理学,神经科学提出新的问题和方向。如验证信息处理的通路,神经网络的局部回路,甚至可以完成一些在活体上无法完成的实验等。
目前国内外已有一些模拟视网膜结构与功能的模型,按其最终的目的和面向的领域来看,大致分三类。第一类：面向神经科学。第二类：面向计算机科学应用。第三类：面向电子工程硬件实现。这些模型或局限于部分的神经回路和信息通路而没有展现视网膜的整体特性,或仅借鉴视网膜模型的结构框架和部分特性而跳过了大量的生理细节。总体上来说,当前已有的视网膜模型与真实生理情况相比,不仅在细胞的类型和数量上还有所欠缺,在视网膜的功能和结构模拟上的工作也不充分,更重要的是他们的信息处理过程和真实的信息处理过程还有很大差距。在目前的情况下,利用已有的神经生物学、解剖学,电生理学等学科的数据和结果,建立一个高度逼真的视网膜计算模型,是一种有益的尝试。
Biologist Rosen have said:"from evolution point of view, physical system is the best encyclopedia for human to solve complex problems." Among all chapters of the encyclopedia, the eye is definitely one of the most brilliant one, which has astonishing powerful function, intricate connections and structure, highly precise control mechanism, even Darwin was shocked by eye and praised it as "extremely perfect and complex organ". Our work is inspired by the most important part of the eye--retina, which is in charge of receiving and processing visual signals.
Long time ago, people had recognized the retina structure is really complex, however, some basic but essential questions, such as how many layers of retina can be divided into, how many types of cells are included in different layer, what is the number of each type of cell and how they distribute, are gradually clear in recent year with the development of anatomy, electrophysiology, cell morphology and so on. Based on these data, we simulated a highly realistic retina model. A variety of experiments can be designed based on the model, such as object detection probability of retina, object representation efficiency of retina, the retina balance property under multiple constrains(such as accuracy, real-time, energy consumption, computation complexity, etc.) and so on.
This model presents a new direction or a new perspective for some difficult problems not well resolved in traditional way. For example, the design of artificial retina chips. Because of limited size, the hardware complexity of artificial retina can not be too high; Because of heat issue, energy consumption can not be too much; Because of real-time response requirements, signal processing must be fast and accurate. It is very hard to balance these constraints which conflict with each other. However, these tasks are exactly what the retina usually does, and it does really well. This inspires us that if the balance mechanism can be introduced into artificial retina design, it would be very helpful. And also, this model can be used for real-time scene representation and image processing. Retina divides visual signals into color pathways, brightness pathways, motion pathways and some other pathways to form internal representation for further processing, which dramatically improves the efficiency of representation and meanwhile significantly reduces the computation complexity. This method may provide a new direction for real-time scene representation and image processing. Meanwhile, since the model is highly close to the real retina, it can test physical experiment to verity the physical assumptions, and future to raise new questions or directions to physiology, neuroscience and other subjects.
Actually, there have been some simulation models on retina. According to their fields and goals, these models can generally be divided into three categories. The first category is mainly for neuroscience. The second category may focus on computer science applications. The third category may face to micro electronic hardware design. These models are often too detailed to demonstrate retina nature as a whole, or often too general to include key parts and essential features, and worse, the information processing process of these models are far from what the real retina does. The model we presented not only simulates a variety of retina cells and their distribution, but also simulates intricate connections and structure of the retina, not only is loyal to the physical processing details, but also is easy to demonstrate the overall features of the retina. We believe this would be a beneficial attempt and provide a new perspective of thinking.
很早之前,人们就已认识到了视网膜的复杂,但具体复杂到什么程度,视网膜可分为哪些层,每层有哪些种细胞,每种细胞的数量和分布如何,则是随着近几年来解剖学,电生理学,细胞形态学等学科的发展而逐渐清晰起来的。基于这些数据,我们建立了一个高度逼真的视网膜计算模型,此模型不仅模拟了视网膜的多种细胞类型和分布,还拟合了视网膜复杂的层次连接和结构,既忠于生理上的结构和处理过程又在整体上展现了视网膜的多个信息处理通路。在模型基础上可以设计很多实验,比如视网膜物体检测概率实验,视网膜物体表征效率实验,视网膜在多条件限定下(如准确性,实时性,能耗,计算复杂度等)的均衡性分析实验等等。
有了这个高度逼真的视网膜计算模型,可以为当前很多棘手的问题提供解决的方向。如人工视网膜芯片的设计。人工视网膜因为尺寸有限,所以硬件复杂度不能太高；因为发热问题,所以能耗不能太大；因为要满足实时的反应要求,所以信号处理需要快速且准确。这些约束条件相互冲突,很难平衡。但这些工作正是视网膜平时所做的事情。研究视网膜在多条件下的平衡性质,对人工视网膜的设计具有巨大的指导意义。此模型也可用于实时的场景表征和图像处理。视网膜将信号分为颜色,亮度,运动等不同的通路形成内部表征再进行进一步处理,这在提高表征的效率的同时大幅降低了计算的复杂度,为实时图像的表征和处理提供了一个新的方向。同时,由于模型高度逼真于真实的视网膜,也可以进行生理实验,验证生理上的假设和推测,为生理学,神经科学提出新的问题和方向。如验证信息处理的通路,神经网络的局部回路,甚至可以完成一些在活体上无法完成的实验等。
目前国内外已有一些模拟视网膜结构与功能的模型,按其最终的目的和面向的领域来看,大致分三类。第一类：面向神经科学。第二类：面向计算机科学应用。第三类：面向电子工程硬件实现。这些模型或局限于部分的神经回路和信息通路而没有展现视网膜的整体特性,或仅借鉴视网膜模型的结构框架和部分特性而跳过了大量的生理细节。总体上来说,当前已有的视网膜模型与真实生理情况相比,不仅在细胞的类型和数量上还有所欠缺,在视网膜的功能和结构模拟上的工作也不充分,更重要的是他们的信息处理过程和真实的信息处理过程还有很大差距。在目前的情况下,利用已有的神经生物学、解剖学,电生理学等学科的数据和结果,建立一个高度逼真的视网膜计算模型,是一种有益的尝试。
Biologist Rosen have said:"from evolution point of view, physical system is the best encyclopedia for human to solve complex problems." Among all chapters of the encyclopedia, the eye is definitely one of the most brilliant one, which has astonishing powerful function, intricate connections and structure, highly precise control mechanism, even Darwin was shocked by eye and praised it as "extremely perfect and complex organ". Our work is inspired by the most important part of the eye--retina, which is in charge of receiving and processing visual signals.
Long time ago, people had recognized the retina structure is really complex, however, some basic but essential questions, such as how many layers of retina can be divided into, how many types of cells are included in different layer, what is the number of each type of cell and how they distribute, are gradually clear in recent year with the development of anatomy, electrophysiology, cell morphology and so on. Based on these data, we simulated a highly realistic retina model. A variety of experiments can be designed based on the model, such as object detection probability of retina, object representation efficiency of retina, the retina balance property under multiple constrains(such as accuracy, real-time, energy consumption, computation complexity, etc.) and so on.
This model presents a new direction or a new perspective for some difficult problems not well resolved in traditional way. For example, the design of artificial retina chips. Because of limited size, the hardware complexity of artificial retina can not be too high; Because of heat issue, energy consumption can not be too much; Because of real-time response requirements, signal processing must be fast and accurate. It is very hard to balance these constraints which conflict with each other. However, these tasks are exactly what the retina usually does, and it does really well. This inspires us that if the balance mechanism can be introduced into artificial retina design, it would be very helpful. And also, this model can be used for real-time scene representation and image processing. Retina divides visual signals into color pathways, brightness pathways, motion pathways and some other pathways to form internal representation for further processing, which dramatically improves the efficiency of representation and meanwhile significantly reduces the computation complexity. This method may provide a new direction for real-time scene representation and image processing. Meanwhile, since the model is highly close to the real retina, it can test physical experiment to verity the physical assumptions, and future to raise new questions or directions to physiology, neuroscience and other subjects.
Actually, there have been some simulation models on retina. According to their fields and goals, these models can generally be divided into three categories. The first category is mainly for neuroscience. The second category may focus on computer science applications. The third category may face to micro electronic hardware design. These models are often too detailed to demonstrate retina nature as a whole, or often too general to include key parts and essential features, and worse, the information processing process of these models are far from what the real retina does. The model we presented not only simulates a variety of retina cells and their distribution, but also simulates intricate connections and structure of the retina, not only is loyal to the physical processing details, but also is easy to demonstrate the overall features of the retina. We believe this would be a beneficial attempt and provide a new perspective of thinking.
引文
[1]Samir Shah, Martin D. Visual information processing in primate cone pathways. LA model. IEEE Transactions On Systems, Man, And Cybernetics-Part B: Cybernetics.1996; 26(2); 259-274
[2]Samir Shah, Martin D. Visual information processing in primate cone pathways. Ⅱ. Experiments. IEEE Transactions On Systems, Man, And Cybernetics-Part B: Cybernetics.1996; 26(2); 275-289
[3]Rodrigo Publio, Rodrigo F. Oliveira, Antonio C. Roque.A realistic model of rod photoreceptor for use in a retina network model. Neurocomputing 69 (2006) 1020-1024
[4]Alexander Casti,Fernand Hayot,Youping Xiao,Ehud Kaplan. A simple model of retina-LGN transmission.J Comput Neurosci (2008) 24:235-252
[5]Felice A. Dunn,Thuy Doan,Alapakkam P. Sampath,Fred Rieke. Controlling the Gain of Rod-Mediated Signals in the Mammalian Retina. The Journal of Neuroscience, April 12, (2006)26(15):3959-3970
[6]Peter H. Schiller, Warren M. Slocum, Veronica S. Weiner. How the parallel channels of the retina contribute to depth processing. European Journal of Neuroscience, (2007)26,1307-1321
[7]Botond Roska, Alyosha Molnar, and Frank S. Werblin. Parallel Processing in Retinal Ganglion Cells:How Integration of Space-Time Patterns of Excitation and Inhibition Form the Spiking Output. J Neurophysiol (2006)95:3810-3822
[8]Wang-Qiang Niu, Jing-Qi Yuan. Recurrent network simulations of two types of non-concentric retinal ganglion cells. Neurocomputing 70 (2007) 2576-2580
[9]Wang-Qiang Niu, Jing-Qi Yuan. A multi-subunit spatiotemporal model of local edge detector cells in the cat retina.Neurocomputing,2008
[10]Qiling Tang, Nong Sang,and Tianxu Zhang. A Neural Network Model for Extraction of Salient Contours. ISNN 2005, LNCS 3497, pp.316-320,2005.
[11]Rong Lu, Yi Shen, and Qiang Wang. Edge Detection Based on Early Vision Model Incorporating Improved Directional Median Filtering. Instrumentation and Measurement Technology Conference,2004. IMTC 04. Proceedings of the 21st IEEE.
[12]Giorgio Cucurachi, Guido Tascini, Francesco Piazza. Neural Network for Region Detection. Proceedings of the 9th International Conference on Image Analysis and Processing-Volume Ⅱ. Pages:228-237
[13]Reddick, WE., Glass, J.O., Cook, E.N., Elkin, T.D., Deaton, R.J. Automated segmentation and classification of multispectral magnetic resonance images of brain using artificial neural networks. Medical Imaging, IEEE Transactions on Volume 16, Issue 6, Dec.1997 Page(s):911-918
[14]Lon Risinger,Khosrow Kaikhah .Motion detection and object tracking with discrete leaky integrate-and-fire neurons. Appl Intell,2007
[15]Lee, J-W., Chae, S-P., Kim, M-N., Kim, S-Y, Cho, J-H. A moving detectable retina model considering the mechanism of an amacrine cell for vision. Industrial Electronics,2001. Proceedings. ISIE 2001. IEEE International Symposium on Volume 1, Issue,2001 Page(s):106-109 vol.1
[16]Naohiro Ishii, Toshinori Deguchi, and Hiroshi Sasaki.Parallel Processing for Movement Detection in Neural Networks with Nonlinear Functions. Intelligent Data Engineering and Automated Learning-IDEAL 2004,5th International Conference, Exeter, UK, August 25-27,2004, Proceedings vol.3177, pp.626-633
[17]Kazuhiro Shimonomura, Takayuki Kushima, Tetsuya Yagi. Binocular robot vision emulating disparity computation in the primary visual cortex. Neural Networks 21(2008)331-340
[18]Behrend, M. R., A. K. Ahuja, M. S. Humayun, J. D. Weiland, and R. H. Chow. 2009. "Selective Labeling of Retinal Ganglion Cells with Calcium Indicators by Retrograde Loading In Vitro," Journal of Neuroscience Methods 179(2),166-72.
[19]Caspi, A., J. D. Dorn, K. H. McClure, M. S. Humayun, R. J. Greenberg, and M. J. McMahon. 2009. "Feasibility Study of a Retinal Prosthesis:Spatial Vision with a 16-Electrode Implant," Archives of Ophthalmology 127(4),398-401.
[20]Ahuja A. K., M. R. Behrend, M. Kuroda, M. S. Humayun, and J. D. Weiland. 2008. "An In Vitro Model of a Retinal Prosthesis," IEEE Transactions on Biomedical Engineering 55(6),1744-53.
[21]Wu, L., Z. Yang, E. Basham, and W. Liu.2008. "An Efficient Wireless Power Link for High Voltage Retinal Implant." Proceedings of Biomedical Circuits and Systems Conference,101-04.
[22].Christian A. Morillas, Samuel F. Romero, Antonio Mart'nez,Francisco J. Pelayo, Eduardo Ros,Eduardo Fern'andez. A design framework to model retinas. BioSystems 87(2007)156-163
[23]Pelayo F.J, Morillas C, Martinez A, Romero S, Ros E, Pino B:A Reconfigurable Machine to Model the First Stages of the Human Visual Pathway. In "Computacion reconfigurable & FPGAs" E.Boemo et al (Eds). I.S.B.N.:84-600-9928-8; Proc. of the 3rd JCRA conference. Madrid, Spain, September-2003.
[24]Pelayo F.J, Romero S, Morillas C, Martinez A, Ros E, Fernandez E. Translating image sequences into spike patterns for cortical neuro-stimulation. Annual Computational Neuroscience Meeting'2003. Alicante, 2003. Computational Neuroscience Association.
[25]Pelayo F.J, Romero S, Morillas C, Ros E, Fernandez E. Cortical visual neuroprostheses for the blind:retina-like software/hardware preprocesor. IEEE EMBS Conference on Neural Engineering Capri, Italia 2003.IEEE EMBS. ISBN: 0-7803-7819-9.
[26]R.A. Normann, D.J. Warren, J. Ammermuller, E. Fernandez, S.Guillory, High-resolution spatio-temporal mapping of visual path-ways using multi-electrode arrays, Vis. Res.41 (2001) 1261-1275
[27]Jost B. Jonas, Ulrike Schneider, and Gottfried O.H. Naumann,1992. Count and density of human retinal photoreceptors. Graefe's Arch Clin Exp Ophthalmol (1992)230:505-510
[28]Sjostrand J, Olsson V, Popovic Z, Conradi N. Quantitative estimations of foveal and extra-foveal retinal circuitry in humans. Vision Res.1999 Sep;39(18):2987-98.
[29]Aggarwal P, Nag TC, Wadhwa S. Age-related decrease in rod bipolar cell density of the human retina:an immunohistochemical study. J Biosci.2007 Mar; 32(2):293-8
[30]Strettoi E, Raviola E, Dacheux RF.Synaptic connections of the narrow-field, bistratified rod amacrine cell (All) in the rabbit retina. J Comp Neurol.1992 Nov 8;325(2):152-68.
[31]E. Brady Trexler, Wei Li, and Stephen C. Massey. Simultaneous Contribution of Two Rod Pathways to All Amacrine and Cone Bipolar Cell Light Responses.J Neurophysiol.November 3,2004.1476-1485
[32]Peter Sterling, Michael A. Freed,Robert G. Smith. Architecture of Rod and Cone Circuits to the On-beta Ganglion Cell.The Journal of Neuroscience, February 1988, 8(2):623-642
[33]Nelson R, Kolb H, Freed MA.OFF-alpha and OFF-beta ganglion cells in cat retina. I:Intracellular electrophysiology and HRP stains. J Comp Neurol.1993 Mar 1;329(1):68-84.
[34]Kolb H, Nelson R.OFF-alpha and OFF-beta ganglion cells in cat retina:Ⅱ. Neural circuitry as revealed by electron microscopy of HRP stains. J Comp Neurol. 1993 Mar 1;329(1):85-110.
[35]Harman A, Abrahams B, Moore S, Hoskins R. Neuronal density in the human retinal ganglion cell layer from 16-77 years. Anat Rec.2000 Oct 1;260(2):124-31.
[36]Dacey DM. The mosaic of midget ganglion cells in the human retina. J Neurosci. 1993 Dec;13(12):5334-55.
[37]D M Dacey, M R Petersen. Dendritic field size and morphology of midget and parasol ganglion cells of the human retina. Proc Natl Acad Sci U S A.1992 October 15; 89(20):9666-9670.
[38]Akiko Iwaizumi,Ryoko Futami,Shin'ichiro Kanoh,Jiro Gyoba. Characteristics of human luminance discrimination and modeling a neural network based on the response properties of the visual cortex. Biol Cybern (2006) 94:381-392
[39]Jae-Sung Kong, Sang-Heon Kim, Jang-Kyoo Shin, and Minho Lee. An Artificial Retina Chip Using Switch-Selective Resistive Network for Intelligent Sensor Systems. ICIC 2006,702-710
[40]Zrenner E. Will retinal implants restore vision? Science,2002,295:1022-1025.
[41]Humayun MS, Propst R, de Juan E, et al. Bipolar surface electrical stimulation of the vertebrate retina. Arch Ophthalmol,1994,112:110-116.
[42]Narayaun MV, Rizzo JF, Edell D,et al. Development of a silicon retinal implant.. Invest Ophthalmol VisSci,1994,38:S1380
[43]Sicard, G. An adaptive bio-inspired analog silicon retina. In:Proceedings of the 25th European Solid-State Circuits Conference,1999; 306-309
[44].Visual Prosthesis and Ophthalmic Devices:Circuit Designs That Model the Properties of the Outer and Inner Retina. Humana Press,2007,135-158
[45].Neural Engineering:Retinal Bioengineering. Springer US,2005,421-484
[46]D. S. Wills, J. M. Baker, H. H. Cat, S. M. Chai, L. Codrescu, J. L. Cruz-Rivera, J. C. Eble, A. Gentile, M. A. Hopper, W. S. Lacy, A. Lopez-Lagunas, P. May, S. Smith, T. Taha, Processing Architectures for Smart Pixel Systems, IEEE Journal of Selected Topics in Quantum Electronics, vol.2 (1),1996,24-34
[47]S.Mertoguno, N.Bourbakis. A digital retina-like low level vision processor. Digital Object Identifier; 2003; 33(5); 782-788
[48]J Herault. A model of colour processing in the retina of vertebrates from photoreceptors to colour opposition and colour constancy phenomena. Neurocomputing.1996; 12; 113-129
[49]Andersen, J.D. Methods for modeling the first layers of the retina. In: Neuroinformatics and Neurocomputers. IEEE,1992; 1; 179-186
[50]Lee, J.-W. A moving detectable retina model considering the mechanism of an amacrine cell for vision. In:Proceedings Industrial Electronics. IEEE,2001; 1; 106-109
[51]Qiu Fangtu, Li Chaoyi. Mathematical simulation of disinhibitory properties of concentric receptive field. ACTA biophysica sinca.2000; 16(2); 214-220
[52]Li Zang, Qiu Zhi -cheng. A new computational model of retinal ganglion cell receptive fields Ⅰ:a model of ganglion cell receptive fields with extended disinhibitory area. ACTA biophysica sinca.2000; 16(2).288-295
[53]Li Zang, Qiu Zhi -cheng. A new computational model of retinal ganglion cell receptive fields Ⅱ:modeling center-surround interactions in orientation selectivity of a ganglion cell receptive field with extended disinhibitory area. ACTA biophysica sinca.2000;16(2);296-302
[54]Kareem A. Zaghloul, Kwabena Boahen. Optic nerve signals in a neuromorphic chip Ⅱ:test and results. Transactions on Biomedical Engineering; 2004; 51(2); 667-675
[55]Sicard, G. An adaptive bio-inspired analog silicon retina.In:Proceedings of the 25th European Solid-State Circuits Conference,1999; 306-309
[56]Carmona, R. A CMOS analog parallel array processor chip with programmable dynamics for early vision tasks. In:Proceedings of the 28th European Solid-State Circuits Conference,2002; 371-374
[57]Joyce Tombran-Tink, Colin J. Barnstable, Joseph F. Rizzo Ⅲ,2007. Visual Prosthesis and Ophthalmic Devices. Humana Press.135-158
[58]KOLB Helga. How the retina works. American scientist ISSN 0003-0996 CODEN AMSCAC.2003, vol.91, pp.28-35
[59]Sumitha Balasuriya, Paul Siebert. A Biologically Inspired Computational Vision Front-end based on a Self-Organised Pseudo-Randomly Tessellated Artificial Retina. IJCNN,2005
[60]寿天德.《视觉信息处理的脑机制》[M].上海科技教育出版社1997
[61]Francis Crick.. The astonishing hypothesis:the scientific search for the soul. London:Simon & Schuster,1994.
[62]Smith,R. G., M. A. Freed, and P. Sterling (1986) Microcircuitry of the dark and light adapted states in the cat retina: Functional architecture of the rod-cone network. J.Neurosci.6:3505-3517.
[63]曹洋,顾凡及视网膜神经节细胞非经典感受野及其方向倾向性的模型研究,复旦学报(自然科学版),2005,44(4)：524-527
[64]戴荣平,董方田 人工眼的研究现状及发展前景 中华眼科杂志2003年4月 第39卷第4期,254—256
[65]Saleem,R.A., Walter,M.A.. The complexities of ocular genetics.2002. Clin. Genet.61:79-88.
[66]Walter,P., Szurman,P., Vobig,M., Berk,H., Ludtke-Handjery,H.C., Richter,H., Deng, Mittermayer,C., Heimann, K.,and Sellhaus, B. Successful long-tern implantation of inactive epiretinal microelectrode arrays in rabbits.1999. Retina 19:546-552.
[67]Walter,P., and Heimann,K. Evoked cortical potentials after electrical stimulation of the inner retina in rabbits.2000. Graefe's Arch. Clin. Exp. Ophthalmol. 238:315-318.
[68]Wang,L., Kondo,M., and Bill,A. Glucose metabolism in cat outer retina. 1997.Invest. Ophthalmol. Visual Sci.38:48-55.
[69]Wangsa-Wirawan,N., and Linsenmeier,R.A. Retinal oxygen:Fundamental and clinical aspects.2003. Arch. Ophthalmol.121:547-557.
[70]Botond Roska, Alyosha Molnar and Frank S. Werblin. Parallel Processing in Retinal Ganglion Cells:How Integration of Space-Time Patterns of Excitation and Inhibition Form the Spiking Output. J Neurophysiol 95:3810-3822,2006.
[71]Ahnelt P, Ammermuller J, Pelayo F, Bongard M, Palomar D, Piedade M, Ferrandez JM, Borg-Graham L, Fernandez E. Neuroscientific basis for the design and development of a bioinspired visual processing front-end. IFMBE Proceedings, Vienna 2002 3(2) pp 1692-1693.
[72]Bongard M, Ammermuller J, Climent R, Ferrandez JM, Normann RA, Ahnelt P, Belmonte J, Fernandez E. In vivo massive parallel recordings of human retinal ganglion cell populations. Ophthal Res 34/S1:363
[73]Bongard M, Ferrandez JM, Bolea JA, Ammermuller J, Fernandez E. Extracting common features from the responses of different ganglion cell populations. Ophthal Res 34/S1:363.
[74]Ferrandez JM, Bongard M, Garcia de Quiros F, Bolea JA, Ammermuller J, Normann RA, Fernandez E. Decoding the population responses of retinal ganglion cells using information theory. in Mira J, Sanchez-Andres JV (eds) Engeneering applications of bio-inspired artificial neural networks. Lecture Notes in Computer Science, Springer Verlag Berlin.
[75]Bonomini P, Bongard M, Ferrandez J.M, Pelayo F, Fernandez E. Population coding in simultaneously recorded retinal ganglion cells. 5th International Workshop on Neural Coding Aulla, Italia 2003.
[76]E. Fernandez; P. Ahnelt; P. Rabischong, C. Botella;F. Garcia-de Quiros; P. Bonomini; C. Marin; R. Climent;J.M. Tormos; and R.A. Normann, "Towards a Cortical Visual Neuroprosthesis for the Blind",. EMBEC 2002
[77]M.Greschner, M. Bongard, P.Rujan and J. Ammermuller,"Retinal ganglion cell synchronization by fixational eye movementsimproves feature estimation. Nature Neuroscience, vol 5, pp 341-345.
[78]Andreu E, Fernandez E, Louis E, Ortega G, Sanchez-Andres JV. Effects on fluctuations on electrical properties of gap-juntion connected cells.Neurosci. Lett.323: 21-24.2002
[79]Fernandez E., Alfaro A., Tormos JM, Climent R., Martinez M., Vilanova H., Walsh V, Pascual-Leone, A. Mapping of the human visual cortex using image-guided transcranial magnetic stimulation. Brain Res. Protocols.10:115-124.2002.
[80]W.H. Dobelle, Artificial vision for the blind by connecting a camera to the visual cortex, ASAIO, Int. J. Artif. Organs (2000) 3-9.
[81]Basinger, B. C., A. P. Rowley, K. Chen, M. S. Humayun, J. D. Weiland.2009. "Finite Element Modeling of Retinal Prosthesis Mechanics," Journal of Neural Engineering 6(5),55006.
[2]Samir Shah, Martin D. Visual information processing in primate cone pathways. Ⅱ. Experiments. IEEE Transactions On Systems, Man, And Cybernetics-Part B: Cybernetics.1996; 26(2); 275-289
[3]Rodrigo Publio, Rodrigo F. Oliveira, Antonio C. Roque.A realistic model of rod photoreceptor for use in a retina network model. Neurocomputing 69 (2006) 1020-1024
[4]Alexander Casti,Fernand Hayot,Youping Xiao,Ehud Kaplan. A simple model of retina-LGN transmission.J Comput Neurosci (2008) 24:235-252
[5]Felice A. Dunn,Thuy Doan,Alapakkam P. Sampath,Fred Rieke. Controlling the Gain of Rod-Mediated Signals in the Mammalian Retina. The Journal of Neuroscience, April 12, (2006)26(15):3959-3970
[6]Peter H. Schiller, Warren M. Slocum, Veronica S. Weiner. How the parallel channels of the retina contribute to depth processing. European Journal of Neuroscience, (2007)26,1307-1321
[7]Botond Roska, Alyosha Molnar, and Frank S. Werblin. Parallel Processing in Retinal Ganglion Cells:How Integration of Space-Time Patterns of Excitation and Inhibition Form the Spiking Output. J Neurophysiol (2006)95:3810-3822
[8]Wang-Qiang Niu, Jing-Qi Yuan. Recurrent network simulations of two types of non-concentric retinal ganglion cells. Neurocomputing 70 (2007) 2576-2580
[9]Wang-Qiang Niu, Jing-Qi Yuan. A multi-subunit spatiotemporal model of local edge detector cells in the cat retina.Neurocomputing,2008
[10]Qiling Tang, Nong Sang,and Tianxu Zhang. A Neural Network Model for Extraction of Salient Contours. ISNN 2005, LNCS 3497, pp.316-320,2005.
[11]Rong Lu, Yi Shen, and Qiang Wang. Edge Detection Based on Early Vision Model Incorporating Improved Directional Median Filtering. Instrumentation and Measurement Technology Conference,2004. IMTC 04. Proceedings of the 21st IEEE.
[12]Giorgio Cucurachi, Guido Tascini, Francesco Piazza. Neural Network for Region Detection. Proceedings of the 9th International Conference on Image Analysis and Processing-Volume Ⅱ. Pages:228-237
[13]Reddick, WE., Glass, J.O., Cook, E.N., Elkin, T.D., Deaton, R.J. Automated segmentation and classification of multispectral magnetic resonance images of brain using artificial neural networks. Medical Imaging, IEEE Transactions on Volume 16, Issue 6, Dec.1997 Page(s):911-918
[14]Lon Risinger,Khosrow Kaikhah .Motion detection and object tracking with discrete leaky integrate-and-fire neurons. Appl Intell,2007
[15]Lee, J-W., Chae, S-P., Kim, M-N., Kim, S-Y, Cho, J-H. A moving detectable retina model considering the mechanism of an amacrine cell for vision. Industrial Electronics,2001. Proceedings. ISIE 2001. IEEE International Symposium on Volume 1, Issue,2001 Page(s):106-109 vol.1
[16]Naohiro Ishii, Toshinori Deguchi, and Hiroshi Sasaki.Parallel Processing for Movement Detection in Neural Networks with Nonlinear Functions. Intelligent Data Engineering and Automated Learning-IDEAL 2004,5th International Conference, Exeter, UK, August 25-27,2004, Proceedings vol.3177, pp.626-633
[17]Kazuhiro Shimonomura, Takayuki Kushima, Tetsuya Yagi. Binocular robot vision emulating disparity computation in the primary visual cortex. Neural Networks 21(2008)331-340
[18]Behrend, M. R., A. K. Ahuja, M. S. Humayun, J. D. Weiland, and R. H. Chow. 2009. "Selective Labeling of Retinal Ganglion Cells with Calcium Indicators by Retrograde Loading In Vitro," Journal of Neuroscience Methods 179(2),166-72.
[19]Caspi, A., J. D. Dorn, K. H. McClure, M. S. Humayun, R. J. Greenberg, and M. J. McMahon. 2009. "Feasibility Study of a Retinal Prosthesis:Spatial Vision with a 16-Electrode Implant," Archives of Ophthalmology 127(4),398-401.
[20]Ahuja A. K., M. R. Behrend, M. Kuroda, M. S. Humayun, and J. D. Weiland. 2008. "An In Vitro Model of a Retinal Prosthesis," IEEE Transactions on Biomedical Engineering 55(6),1744-53.
[21]Wu, L., Z. Yang, E. Basham, and W. Liu.2008. "An Efficient Wireless Power Link for High Voltage Retinal Implant." Proceedings of Biomedical Circuits and Systems Conference,101-04.
[22].Christian A. Morillas, Samuel F. Romero, Antonio Mart'nez,Francisco J. Pelayo, Eduardo Ros,Eduardo Fern'andez. A design framework to model retinas. BioSystems 87(2007)156-163
[23]Pelayo F.J, Morillas C, Martinez A, Romero S, Ros E, Pino B:A Reconfigurable Machine to Model the First Stages of the Human Visual Pathway. In "Computacion reconfigurable & FPGAs" E.Boemo et al (Eds). I.S.B.N.:84-600-9928-8; Proc. of the 3rd JCRA conference. Madrid, Spain, September-2003.
[24]Pelayo F.J, Romero S, Morillas C, Martinez A, Ros E, Fernandez E. Translating image sequences into spike patterns for cortical neuro-stimulation. Annual Computational Neuroscience Meeting'2003. Alicante, 2003. Computational Neuroscience Association.
[25]Pelayo F.J, Romero S, Morillas C, Ros E, Fernandez E. Cortical visual neuroprostheses for the blind:retina-like software/hardware preprocesor. IEEE EMBS Conference on Neural Engineering Capri, Italia 2003.IEEE EMBS. ISBN: 0-7803-7819-9.
[26]R.A. Normann, D.J. Warren, J. Ammermuller, E. Fernandez, S.Guillory, High-resolution spatio-temporal mapping of visual path-ways using multi-electrode arrays, Vis. Res.41 (2001) 1261-1275
[27]Jost B. Jonas, Ulrike Schneider, and Gottfried O.H. Naumann,1992. Count and density of human retinal photoreceptors. Graefe's Arch Clin Exp Ophthalmol (1992)230:505-510
[28]Sjostrand J, Olsson V, Popovic Z, Conradi N. Quantitative estimations of foveal and extra-foveal retinal circuitry in humans. Vision Res.1999 Sep;39(18):2987-98.
[29]Aggarwal P, Nag TC, Wadhwa S. Age-related decrease in rod bipolar cell density of the human retina:an immunohistochemical study. J Biosci.2007 Mar; 32(2):293-8
[30]Strettoi E, Raviola E, Dacheux RF.Synaptic connections of the narrow-field, bistratified rod amacrine cell (All) in the rabbit retina. J Comp Neurol.1992 Nov 8;325(2):152-68.
[31]E. Brady Trexler, Wei Li, and Stephen C. Massey. Simultaneous Contribution of Two Rod Pathways to All Amacrine and Cone Bipolar Cell Light Responses.J Neurophysiol.November 3,2004.1476-1485
[32]Peter Sterling, Michael A. Freed,Robert G. Smith. Architecture of Rod and Cone Circuits to the On-beta Ganglion Cell.The Journal of Neuroscience, February 1988, 8(2):623-642
[33]Nelson R, Kolb H, Freed MA.OFF-alpha and OFF-beta ganglion cells in cat retina. I:Intracellular electrophysiology and HRP stains. J Comp Neurol.1993 Mar 1;329(1):68-84.
[34]Kolb H, Nelson R.OFF-alpha and OFF-beta ganglion cells in cat retina:Ⅱ. Neural circuitry as revealed by electron microscopy of HRP stains. J Comp Neurol. 1993 Mar 1;329(1):85-110.
[35]Harman A, Abrahams B, Moore S, Hoskins R. Neuronal density in the human retinal ganglion cell layer from 16-77 years. Anat Rec.2000 Oct 1;260(2):124-31.
[36]Dacey DM. The mosaic of midget ganglion cells in the human retina. J Neurosci. 1993 Dec;13(12):5334-55.
[37]D M Dacey, M R Petersen. Dendritic field size and morphology of midget and parasol ganglion cells of the human retina. Proc Natl Acad Sci U S A.1992 October 15; 89(20):9666-9670.
[38]Akiko Iwaizumi,Ryoko Futami,Shin'ichiro Kanoh,Jiro Gyoba. Characteristics of human luminance discrimination and modeling a neural network based on the response properties of the visual cortex. Biol Cybern (2006) 94:381-392
[39]Jae-Sung Kong, Sang-Heon Kim, Jang-Kyoo Shin, and Minho Lee. An Artificial Retina Chip Using Switch-Selective Resistive Network for Intelligent Sensor Systems. ICIC 2006,702-710
[40]Zrenner E. Will retinal implants restore vision? Science,2002,295:1022-1025.
[41]Humayun MS, Propst R, de Juan E, et al. Bipolar surface electrical stimulation of the vertebrate retina. Arch Ophthalmol,1994,112:110-116.
[42]Narayaun MV, Rizzo JF, Edell D,et al. Development of a silicon retinal implant.. Invest Ophthalmol VisSci,1994,38:S1380
[43]Sicard, G. An adaptive bio-inspired analog silicon retina. In:Proceedings of the 25th European Solid-State Circuits Conference,1999; 306-309
[44].Visual Prosthesis and Ophthalmic Devices:Circuit Designs That Model the Properties of the Outer and Inner Retina. Humana Press,2007,135-158
[45].Neural Engineering:Retinal Bioengineering. Springer US,2005,421-484
[46]D. S. Wills, J. M. Baker, H. H. Cat, S. M. Chai, L. Codrescu, J. L. Cruz-Rivera, J. C. Eble, A. Gentile, M. A. Hopper, W. S. Lacy, A. Lopez-Lagunas, P. May, S. Smith, T. Taha, Processing Architectures for Smart Pixel Systems, IEEE Journal of Selected Topics in Quantum Electronics, vol.2 (1),1996,24-34
[47]S.Mertoguno, N.Bourbakis. A digital retina-like low level vision processor. Digital Object Identifier; 2003; 33(5); 782-788
[48]J Herault. A model of colour processing in the retina of vertebrates from photoreceptors to colour opposition and colour constancy phenomena. Neurocomputing.1996; 12; 113-129
[49]Andersen, J.D. Methods for modeling the first layers of the retina. In: Neuroinformatics and Neurocomputers. IEEE,1992; 1; 179-186
[50]Lee, J.-W. A moving detectable retina model considering the mechanism of an amacrine cell for vision. In:Proceedings Industrial Electronics. IEEE,2001; 1; 106-109
[51]Qiu Fangtu, Li Chaoyi. Mathematical simulation of disinhibitory properties of concentric receptive field. ACTA biophysica sinca.2000; 16(2); 214-220
[52]Li Zang, Qiu Zhi -cheng. A new computational model of retinal ganglion cell receptive fields Ⅰ:a model of ganglion cell receptive fields with extended disinhibitory area. ACTA biophysica sinca.2000; 16(2).288-295
[53]Li Zang, Qiu Zhi -cheng. A new computational model of retinal ganglion cell receptive fields Ⅱ:modeling center-surround interactions in orientation selectivity of a ganglion cell receptive field with extended disinhibitory area. ACTA biophysica sinca.2000;16(2);296-302
[54]Kareem A. Zaghloul, Kwabena Boahen. Optic nerve signals in a neuromorphic chip Ⅱ:test and results. Transactions on Biomedical Engineering; 2004; 51(2); 667-675
[55]Sicard, G. An adaptive bio-inspired analog silicon retina.In:Proceedings of the 25th European Solid-State Circuits Conference,1999; 306-309
[56]Carmona, R. A CMOS analog parallel array processor chip with programmable dynamics for early vision tasks. In:Proceedings of the 28th European Solid-State Circuits Conference,2002; 371-374
[57]Joyce Tombran-Tink, Colin J. Barnstable, Joseph F. Rizzo Ⅲ,2007. Visual Prosthesis and Ophthalmic Devices. Humana Press.135-158
[58]KOLB Helga. How the retina works. American scientist ISSN 0003-0996 CODEN AMSCAC.2003, vol.91, pp.28-35
[59]Sumitha Balasuriya, Paul Siebert. A Biologically Inspired Computational Vision Front-end based on a Self-Organised Pseudo-Randomly Tessellated Artificial Retina. IJCNN,2005
[60]寿天德.《视觉信息处理的脑机制》[M].上海科技教育出版社1997
[61]Francis Crick.. The astonishing hypothesis:the scientific search for the soul. London:Simon & Schuster,1994.
[62]Smith,R. G., M. A. Freed, and P. Sterling (1986) Microcircuitry of the dark and light adapted states in the cat retina: Functional architecture of the rod-cone network. J.Neurosci.6:3505-3517.
[63]曹洋,顾凡及视网膜神经节细胞非经典感受野及其方向倾向性的模型研究,复旦学报(自然科学版),2005,44(4)：524-527
[64]戴荣平,董方田 人工眼的研究现状及发展前景 中华眼科杂志2003年4月 第39卷第4期,254—256
[65]Saleem,R.A., Walter,M.A.. The complexities of ocular genetics.2002. Clin. Genet.61:79-88.
[66]Walter,P., Szurman,P., Vobig,M., Berk,H., Ludtke-Handjery,H.C., Richter,H., Deng, Mittermayer,C., Heimann, K.,and Sellhaus, B. Successful long-tern implantation of inactive epiretinal microelectrode arrays in rabbits.1999. Retina 19:546-552.
[67]Walter,P., and Heimann,K. Evoked cortical potentials after electrical stimulation of the inner retina in rabbits.2000. Graefe's Arch. Clin. Exp. Ophthalmol. 238:315-318.
[68]Wang,L., Kondo,M., and Bill,A. Glucose metabolism in cat outer retina. 1997.Invest. Ophthalmol. Visual Sci.38:48-55.
[69]Wangsa-Wirawan,N., and Linsenmeier,R.A. Retinal oxygen:Fundamental and clinical aspects.2003. Arch. Ophthalmol.121:547-557.
[70]Botond Roska, Alyosha Molnar and Frank S. Werblin. Parallel Processing in Retinal Ganglion Cells:How Integration of Space-Time Patterns of Excitation and Inhibition Form the Spiking Output. J Neurophysiol 95:3810-3822,2006.
[71]Ahnelt P, Ammermuller J, Pelayo F, Bongard M, Palomar D, Piedade M, Ferrandez JM, Borg-Graham L, Fernandez E. Neuroscientific basis for the design and development of a bioinspired visual processing front-end. IFMBE Proceedings, Vienna 2002 3(2) pp 1692-1693.
[72]Bongard M, Ammermuller J, Climent R, Ferrandez JM, Normann RA, Ahnelt P, Belmonte J, Fernandez E. In vivo massive parallel recordings of human retinal ganglion cell populations. Ophthal Res 34/S1:363
[73]Bongard M, Ferrandez JM, Bolea JA, Ammermuller J, Fernandez E. Extracting common features from the responses of different ganglion cell populations. Ophthal Res 34/S1:363.
[74]Ferrandez JM, Bongard M, Garcia de Quiros F, Bolea JA, Ammermuller J, Normann RA, Fernandez E. Decoding the population responses of retinal ganglion cells using information theory. in Mira J, Sanchez-Andres JV (eds) Engeneering applications of bio-inspired artificial neural networks. Lecture Notes in Computer Science, Springer Verlag Berlin.
[75]Bonomini P, Bongard M, Ferrandez J.M, Pelayo F, Fernandez E. Population coding in simultaneously recorded retinal ganglion cells. 5th International Workshop on Neural Coding Aulla, Italia 2003.
[76]E. Fernandez; P. Ahnelt; P. Rabischong, C. Botella;F. Garcia-de Quiros; P. Bonomini; C. Marin; R. Climent;J.M. Tormos; and R.A. Normann, "Towards a Cortical Visual Neuroprosthesis for the Blind",. EMBEC 2002
[77]M.Greschner, M. Bongard, P.Rujan and J. Ammermuller,"Retinal ganglion cell synchronization by fixational eye movementsimproves feature estimation. Nature Neuroscience, vol 5, pp 341-345.
[78]Andreu E, Fernandez E, Louis E, Ortega G, Sanchez-Andres JV. Effects on fluctuations on electrical properties of gap-juntion connected cells.Neurosci. Lett.323: 21-24.2002
[79]Fernandez E., Alfaro A., Tormos JM, Climent R., Martinez M., Vilanova H., Walsh V, Pascual-Leone, A. Mapping of the human visual cortex using image-guided transcranial magnetic stimulation. Brain Res. Protocols.10:115-124.2002.
[80]W.H. Dobelle, Artificial vision for the blind by connecting a camera to the visual cortex, ASAIO, Int. J. Artif. Organs (2000) 3-9.
[81]Basinger, B. C., A. P. Rowley, K. Chen, M. S. Humayun, J. D. Weiland.2009. "Finite Element Modeling of Retinal Prosthesis Mechanics," Journal of Neural Engineering 6(5),55006.