Bayesian statistical MUNE method

Abstract

We have developed a new method of motor unit number estimation (MUNE) for assessing diseases such as amyotrophic lateral sclerosis (ALS). We used data from the whole stimulus-response curve and then performed a Bayesian statistical analysis. The Bayesian method uses mathematical equations that express the basic elements of motor unit activation after electrical stimulation and allows for the sources of variability and uncertainty in this formulation. The Bayesian MUNE method was used to determine the most probable number of motor units in 8 normal subjects, 49 ALS subjects, and 3 subjects with progressive lower motor neuron (LMN) weakness. In normals the number of motor units was calculated to be 75-85 in hand and 40-58 in foot muscles. In ALS subjects the number of motor units per muscle was less than in normal subjects. In 17 ALS subjects and 3 subjects with LMN weakness the median, ulnar, or peroneal nerve was studied on repeated occasions over an average of 189 days (range 63-1,071) and the number of motor units progressively declined, with a half-life ranging from 62-834 days. The results of our MUNE technique were reproducible on replicate studies. A Bayesian statistical MUNE method is a new approach that can be used to study ALS patients serially for assessment and treatment trials.