"Snap on" or Not? A Validation on the Measurement Tool in a Virtual Reality Application

This multi-rater comparison study aims to validate the measurement tool with a "snap" feature option (SNAP ON vs. SNAP OFF), in a virtual reality (VR) application, ImmersiveView v. 2.1, against a conventional software Mimics Innovation Suite v.22 (MIS). It is hypothesized that these measurement tools are equivalent between SNAP ON, and SNAP OFF, and when compared to MIS, in terms of basic linear and angular measurements. Six (6) raters conducted a set of 40 linear and 15 angular measurements using CT scan data of three objects (L-block, hand model, and dry skull) with fiducial markers. Inter-rater repeatability and intra-rater reproducibility were assessed via inter-class coefficient (ICC). Equivalency between each pair of modules (SNAP ON, SNAP OFF, and MIS) was analyzed via Bland-Altman plots and two one-sided t-tests (TOST) procedure. The ICC for intra-rater repeatability yielded 0.999 to 1.000, and inter-rater reproducibility yielded 0.998 to 1.000, which suggests high degree of intra- and inter-rater reliability. The Bland-Altman plots demonstrated that measurements acquired from SNAP ON, SNAP OFF, and MIS were equivalent. The TOST procedure yielded that the measurements through all three modules are equivalent within ± 0.2 mm interval for distance, and ± 0.3° interval for angular measurements. The measurement tool with the "snap" feature in a newly developed VR application (ImmersiveView v.2.1) has been validated through a multi-rater comparison study. In terms of linear and angular measurements, this VR application, whether the "snap" feature was on or off, was equivalent to each other and to the control software (MIS) under the condition of this study. A strong reliability, both intra-rater repeatability and inter-rater reproducibility, has been found.