A self-referential outlier detection method for quantitative motor unit action potential analysis

Abstract Quantitative MUAP analysis is often based on outlier detection, in the case of neurogenic conditions, the finding of MUAPs that are larger than the limit determined from a reference normal population. Such reference data is available from only a few sources and for only a few muscles. It would be desirable if muscles could serve as their own controls. The Henneman size principle determines the order of recruitment, following an exponential distribution of twitch force, motor neurone, motor unit, and MUAP size. Therefore, an outlier could be detected by being too large for the level of recruitment, as judged by its size relative to the other MUAPs. This would improve the sensitivity of detecting neurogenic muscles and obviate the need for external reference data.