All foot models suffer from inherent error resulting from the close proximity of markers. The quantification of marker placement error to understand inherent modeling errors and evaluate techniques of referencing the anatomy is a non trivial problem. A casting method to ensure reproducibility of foot placement and segment orientation was adopted for the current study which examines both the repeatability of placing physical markers (PM) and pointing to anatomical landmarks with a device to create virtual points (VP). Two collaborating sites each investigated three normal adults (n=6 subjects) with two clinicians (n=4 clinicians) performing two data collections sessions on each subject. Two aims were undertaken in this study: 1) to evaluate the efficacy of the casting method, and 2) to evaluate errors in physical marker placement and virtual pointer techniques when identifying foot anatomy. For aim 1, seven physical markers were placed on common landmarks of the foot. Two trials were collected with subject’s feet in the plaster cast. The comparisons of marker locations for these two trials were used to assess the efficacy of casting. The second aim was evaluated by having each clinician repeat the identification of the four key anatomical points in each of two test sessions by applying four PM and by pointing to four VP. This allowed repeatability of PM and VP to be tested within the same clinician. Having multiple clinicians (2) perform the protocol on the same subject allowed for an inter-clinician repeatability measurement.

Positioning and repositioning a subject’s foot while seated using a plaster cast was able to allow repeatability on average of less than 2 mm overall distance. Since the ultimate aim of anatomical identification is not absolute segment location, but rather foot segment angular orientation, these angular measures were assessed in this study. Clinicians were able to reposition markers and point to anatomical landmarks on the same subjects to measure transverse segment attitude to within less than 4° error. Only slightly larger errors were evident when the different clinicians evaluated the same subject (maximum group mean 4.8°). A review of the results indicates the VP techniques are at least as repeatable as PM placement, and with greater familiarity in the VP technique by practicing clinicians, this approach may prove to be superior.