Russell A F, Armiger R S, Vogelstein R J, Bensmaia S J, Etienne-Cummings R
Department of Electrical and Computer Engineering at Johns Hopkins University, Baltimore, MD 21218, USA.
Annu Int Conf IEEE Eng Med Biol Soc. 2009;2009:185-8. doi: 10.1109/IEMBS.2009.5333565.
In order for the functionality of an upper-limb prosthesis to approach that of a real limb it must be able to, accurately and intuitively, convey sensory feedback to the limb user. This paper presents results of the real-time implementation of a 'biofidelic' model that describes mechanotransduction in Slowly Adapting Type 1 (SA1) afferent fibers. The model accurately predicts the timing of action potentials for arbitrary force or displacement stimuli and its output can be used as stimulation times for peripheral nerve stimulation by a neuroprosthetic device. The model performance was verified by comparing the predicted action potential (or spike) outputs against measured spike outputs for different vibratory stimuli. Furthermore experiments were conducted to show that, like real SA1 fibers, the model's spike rate varies according to input pressure and that a periodic 'tapping' stimulus evokes periodic spike outputs.
为了使上肢假肢的功能接近真实肢体，它必须能够准确且直观地向肢体使用者传达感觉反馈。本文展示了一个“生物逼真”模型实时实现的结果，该模型描述了慢适应性1型（SA1）传入纤维中的机械转导。该模型能准确预测任意力或位移刺激下动作电位的时间，其输出可作为神经假体装置对外周神经刺激的刺激时间。通过比较不同振动刺激下预测的动作电位（或尖峰）输出与测量的尖峰输出，验证了模型性能。此外，还进行了实验以表明，与真实的SA1纤维一样，该模型的尖峰频率会根据输入压力而变化，并表明周期性的“轻敲”刺激会引发周期性的尖峰输出。
