Clark F J, Larwood K J, Davis M E, Deffenbacher K A
Department of Physiology and Biophysics, University of Nebraska College of Medicine, Omaha 68198-4575, USA.
Exp Brain Res. 1995;107(1):73-9. doi: 10.1007/BF00228018.
In this paper, we present a method for assessing the exactness of sensing and setting the positions of joints and limbs, using a measure we call target resolution. Target resolution, derived from information theory but ultimately based on variance, estimates the fewest number of discrete, equally spaced targets required within a range to provide the maximum possible information transfer from any target set. We argue that target resolution provides better insight into the exactness of position sense than does the usual measure of accuracy based on mean or constant error. Studies have shown that measures of mean error in setting or indicating positions of joints or limbs exhibit lability; they drift and show considerable sensitivity to factors such as previous positions of the limb and learning. We derive the equation for calculating target resolution and give example resolutions for several joints we have tested. Target resolution often gives a quite different impression of proprioceptive exactness than do measures of accuracy based on mean error.
在本文中,我们提出了一种评估关节和肢体传感准确性以及设置其位置的方法,该方法使用一种我们称为目标分辨率的度量。目标分辨率源自信息论,但最终基于方差,它估计在一个范围内所需的最少离散、等距目标数量,以提供从任何目标集进行最大可能信息传递。我们认为,与基于均值或恒定误差的通常准确性度量相比,目标分辨率能更好地洞察位置感知的准确性。研究表明,在设置或指示关节或肢体位置时,均值误差度量表现出不稳定性;它们会漂移,并且对诸如肢体先前位置和学习等因素表现出相当的敏感性。我们推导了计算目标分辨率的方程,并给出了我们测试过的几个关节的示例分辨率。与基于均值误差的准确性度量相比,目标分辨率常常会给出关于本体感觉准确性的截然不同的印象。