Development of a Technological Platform to Study the Neural Code for Vision

神经科学面临的最大挑战之一是了解神经细胞如何协同作用以创造感知，思想和情感。 在过去的16年里，PI指导了波士顿视网膜植入项目，该项目的目标是开发一种修复术，以恢复盲人的视力。 这项工作的外推强烈表明，即使是一个完美的设计设备也不会成功地产生高质量的视觉。 缺少的因素是如何与中枢神经系统沟通的知识，视网膜是其中的一部分。 神经编码很复杂;对它的特性的深入了解将逐渐发生，而且需要比这笔赠款所涵盖的时间长得多的时间。 PI在该项目这一阶段的目标是开发平台技术，以推进神经编码的研究，并进行学习实验，提供有关视觉系统自然编码的新信息，以及如何提供人工刺激来模拟自然反应。 利用他的研究小组的现有优势，PI将建立设备，使记录和无线传输视网膜或大脑大区域的神经反应成为可能，最终目标是在清醒的动物在测试环境中自由漫游时捕捉这些反应。 计算策略将用于比较通过用光刺激视网膜产生的神经反应与通过以相同几何图案传递的电刺激视网膜产生的神经反应。 学习算法将用于调整电刺激的模式，以模拟自然光诱导的反应。 总的来说，这些研究将为神经编码的特性以及如何模仿大脑的自然反应来创造视觉提供新的见解，并使我们在恢复残疾患者功能的道路上沿着更进一步。更广泛的影响：这项工作将增加视网膜假体有朝一日恢复盲人有用视力的可能性。 数百万患有年龄相关性黄斑变性的患者和另外160万患有视网膜色素变性的患者由于相对选择性的光感受器丧失而失明。 视网膜假体可以通过直接刺激连接眼睛和大脑的神经细胞来恢复这些患者的视力。 只有学会了向大脑传递视觉信息的适当交流策略，这个希望才能实现。 这项研究还将培养工程和生物医学学科的学生，他们将为神经编码，脑机接口和学习科学等新兴领域做出贡献。

One of the great challenges of neuroscience is to understand how nerve cells act synergistically to create perception, thought, and emotion. The PI has for the last 16 years directed the Boston Retinal Implant Project, the goal of which is to develop a prosthesis to restore vision to the blind. An extrapolation of this work strongly suggests that even a perfectly engineered device will not succeed in producing high quality vision. The missing factor is the knowledge of how to communicate with the central nervous system, of which the retina is a part. The neural code is complex; developing insights into its properties will occur gradually, and over a much longer period of time than this grant will cover. The PI's goals for this phase of the project are to develop platform technologies to advance the study of neural coding, and to perform learning experiments that will provide new information about the natural coding of the visual system and how artificial stimuli can be delivered to emulate natural responses. Using the existing strengths of his research group, the PI will build devices that will make it possible to record and wirelessly transmit neural responses from large regions of the retina or brain, with the ultimate goal of capturing these responses in awake animals as they roam freely within a test environment. Computational strategies will be used to compare neural responses that are generated by stimulating the retina with light to those generated by stimulating the retina with electricity delivered in the same geometric patterns. Learning algorithms will be used to adjust the patterns of electrical stimulation to emulate the natural light-induced responses. Collectively, these studies will provide new insights into properties of neural coding and how the natural responses of the brain can be emulated to create vision, and take us a few steps further along the path toward restoration of function to disabled patients.Broader Impacts: This work will increase the probability that a retinal prosthesis will one day restore useful vision to blind patients. Many millions of patients with age-related macular degeneration and another 1.6 million with retinitis pigmentosa suffer from blindness because of a relatively selective loss of photoreceptors. A retinal prosthesis could restore vision to such patients by directly stimulating the nerve cells that connect the eye to the brain. This hope will be realized only if proper communication strategies to deliver visual information to the brain are learned. This research will also train students in engineering and biomedical disciplines who will contribute to the emerging fields of neural coding, brain-machine interfaces and the science of learning.