Numerical Solution of an Identifcation Problem in Electromyography

Titelangaben

Sproll, Tobias ; Schiela, Anton:
Numerical Solution of an Identifcation Problem in Electromyography.
Bayreuth , 2023 . - 21 S.

Volltext

	Format: PDF Name: Sproll_Schiela_NumSolultionIdenficationProblem.pdf Version: Preprint Verfügbar mit der Lizenz Creative Commons BY 4.0: Namensnennung
	Download (625kB)

Abstract

In medical treatment, it can be necessary to identify specific motor units in muscles from measurements on the skin. In [24], we already derived and analyzed an optimization problem suitable for such an identification. This work presents an optimization algorithm that exploits the specific mathematical structure of this problem. Our algorithm is based on a Newton line-search approach in function space with inexact function evaluations due to adaptive numerical quadrature. We present a global convergence result for this method and describe in detail its algorithmic implementation. Finally, we illustrate its practical performance by a numerical study.