Static Precision of Instrumented Pointers for Anatomical Landmark Calibration in CAST-Like Motion Analysis Measurements

TL;DR

This paper studies the precision of instrumented pointers used in motion analysis to improve anatomical landmark calibration accuracy.

Contribution

The first comprehensive characterization of static precision of instrumented pointers in motion capture systems.

Findings

Stationary pointer tip measurements show 0.2 mm variation, while single markers have less than 0.06 mm variation.

Pointer marker proximity affects precision more than pointer geometry.

Modern motion capture systems already provide excellent static precision.

Abstract

The Calibrated Anatomical Systems Technique is an integral part of modern motion analysis systems. However, the calibration of anatomical landmarks is shown to have large variations in intra- and inter-examiner accuracy, which can result in both offset type errors or changes in the characteristics of joint angles and other parameters. This paper is the first instalment in a series of articles aiming to characterize and minimize all of the different factors contributing to these inconsistent calibrations by examining and optimizing the performance of the instrumented pointers used for landmark calibration. A complete characterisation of all aspects of instrumented pointer precision has not been done before.

This paper focuses on examining four different pointers used with an optical OptiTrack motion capture system to establish the expected variability when measuring pointer tip location. Four different pointers were measured at three different locations within the motion capture system’s measurement volume, in a distinct orientation at each of these location.

A single stationary marker can be measured with less than 0.06 mm variation with 95% confidence, whilst the variation of the tip of a stationary pointer is 0.2 mm. If the pointer markers are located closer than what the motion capture system is able to resolve, these variations can more than double, but pointer geometry has limited effect on precision apart from this.

Thanks to improvement of motion capture technology in the last 20 years, static precision is already excellent. Robustness of tracking can likely be improved, but it’s effect on overall pointer precision would be minimal and likely inconsequential.

Figures