Normal goniometric values to guide decision-making in lower-extremity rotational problems using support vector machine techniques

Abstract

Torsional analysis of the lower extremities has become an integral part of the decisionmaking process in treating neuromuscular problems. Solid knowledge of normal development of torsional relationships is essential for treating musculoskeletal problems. As a prerequisite, a normal reference, meaning an objective and quantitative standard of measurement, must be available for comparison prior to making a suitable decision. The aim of this paper is to update normal goniometric values that could be used as a reference for clinical analysis and to discover unknown compensatory mechanisms between the involved segments. A systematic and coherent database made from almost 900 measurements of different parameters has been studied using a process of support vector data description (SVDD) to achieve our outlined targets. The aim is to produce updated decision-making support for clinicians working in the biomechanical exploration of the lower extremities.