Spinal Cord-to-Computer Interface

描述(申请人提供)：由于创伤或疾病造成脊髓损伤后，损伤部位远端的骨骼肌由于神经传导中断而瘫痪。在高度脊髓损伤/损伤中，迫切需要一种可以取代自愿控制的方法，以恢复自我活动、环境控制和计算机访问。在美国，每年大约有15,000例脊髓损伤。这些病例中的大多数幸存下来，并依赖他人满足其基本需求。这一人群的平均预期寿命为40岁。目前的解决方案：“脑-计算机接口”已经被开发出来，可以从不同的大脑皮层提取意志控制信息。通过植入微电极/导线阵列记录单个神经元的活动，并使用先进的信号处理算法进行解释。缺点：经过多年的研究，大脑皮层单棘波记录法仍然存在两个主要问题。首先，记录的神经元数量每天都在变化，因此在每次使用之前需要对信号处理算法进行培训。记录阵列中神经活动的良好电极的数量(产量)很低，由于电极周围神经胶质细胞的生长，单峰信号在一段时间后完全消失。第二，每个神经元提供的信息非常嘈杂。需要对大量神经元进行采样，以获得稳定且可微调的命令信号。这需要在多个脑区植入多个电极阵列。我们的建议：这里提出的替代方法是从损伤部位上方仍然完整的近端脊髓提取意志运动信号，并使用运动束中轴突的群体活动而不是单个棘波。切断的运动神经轴突的远端部分经历了沃勒变性。然而，轴突的近端部分在受伤数年后继续发挥作用，因为它与大脑中的细胞体的连接保持完好。脊髓入路至少有两个重要的优点。首先，由于脊髓在信号路径中与运动装置的距离很近，记录的神经信号将与运动功能强烈耦合。其次，神经记录将更加稳定，因为该方法依赖于群体活动，而不是单个尖峰。

DESCRIPTION (provided by applicant): Following spinal cord damage from trauma or disease, skeletal muscles distal to the point of damage become paralyzed due to disrupted neural conduction. In high-level spinal cord injury/damage, there is a great need for a method that can substitute for the voluntary control in order to regain self-mobility, environmental control, and computer access. Each year about 15,000 spinal cord injuries occur in the US. Majority of these cases survive and depend on others for their basic needs. The average life expectancy of this population is 40 years. Current Solution: 'Brain-Computer Interfaces' have been developed to extract the volitional control information from various brain cortices. Activity of single neurons is recorded with micro electrode/wire arrays implanted and interpreted with advanced signal processing algorithms. Shortcomings: There remain two major problems after many years of research that are inherent to single spike recording method from the cerebral cortex. First, the population of neurons recorded from changes day to day, thus requiring a training session for the signal processing algorithm before each use. The number of good electrodes that record neural activity in an array (yield) is low and the single spike signals are lost completely after sometime due to glial cell growth around the electrode. Second, the information provided by each neuron is very noisy. A very large number of neurons need to be sampled to achieve stable and finely tunable command signals. This requires many electrode arrays implanted in multiple brain areas. Our Proposal: The alternative method proposed here is to extract the volitional motor signals from the proximal spinal cord that is still intact above the site of injury and use the population activity of the axons in the motor tracts rather than single spikes. The distal portions of the severed motor axons go through Wallerian degeneration. However, the proximal part of the axon continues to function years after injury since its connection to the cell body in the brain remains intact. Spinal cord approach has at least two important advantages. First, the recorded neural signals will be strongly coupled to the motor function due to closeness of the spinal cord to the motor apparatus in the signal path. Second, the neural recordings will be much more stable because the method relies on the population activity rather than single spikes.