Digital image segmentation through Bayesian hidden Markov models

Abstract

In motor neurone disease changes in the functional properties of motor units, including the surface voltage, latency, conduction velocity, and response to repetitive stimulation, were investigated. Progression was marked by motor unit loss, increase in the proportion of larger motor unit potentials, and inclusion of motor unit potentials larger than normal in the remaining motor unit population. Even late in the disease, motor unit potentials with a low surface voltage persisted. The relationship between motor unit potentials, surface voltage, and latency, present in control subjects, broke down in motor neurone disease, large motor unit potentials having abnormally long latencies and small motor unit potentials unexpectedly short latencies. Amplitude decrements were more frequent and severe in motor unit potentials at later stages in the disease, particularly in those units with lower surface voltages. In one surviving motor unit potential there was evidence suggestive of functional recovery. The observations point to complex changes in the functional properties of motor units in motor neurone disease.