PC3115: Novel Method for Obtaining Clinical Measurements in a Virtual Environment

Background: 3D modeling and virtual reality (VR) are useful in pre-surgical planning of complex cases1. As use cases in VR expand, the need for more detailed analysis, including measurements, increases. We sought to deploy a novel measurement method that both allowed for scaling in the digital environment while preserving user confidence of the accuracy of the measurement. We chose to utilize Haller index measurements as our comparative value for ease of measurement and ease of bone segmentation. We compare accuracy and precision of measurements made in VR to those in the medical record

Methods: Between 12/2016 and 4/2022, 26 pediatric patients requiring a Nuss procedure had pre-operative CT scans, and Haller indices were recorded by a radiologist. The CT scans were segmented to generate 3D models for inclusion in the VR environment. Similarly, a physical ruler measuring exactly 12 inches was scanned to generate a DICOM 3D model. The resolution of the ruler was insufficient for detailed measurements therefore a digital 3D ruler was created in Solidworks. All three digital models were placed in the same digital environment where scaling applied to each equally. [FIGURE 1] The DICOM segmented ruler was compared to the digitally rendered ruler as the control which was then used as the measuring tool. A second digitally rendered yardstick was generated for measurements greater than 12 inches. As scaling was applied to the clinical model in the digital environment, the same scaling applied to the digital ruler. The DICOM source of both the clinical model and physical ruler transferred confidence to the digital ruler. The transverse and anterior-posterior values were measured in virtual reality using the ruler, and Haller indices were calculated and recorded. [FIGURE 2].

Results: Data was collected, and a t-tailed independent sample t-test was run to compare the similarity of the two groups of indices measured. The average radiologist-measured Haller index was 3.97, compared to the VR-measured Haller index of 4.20. Statistical analysis demonstrated a p-value of 0.56, indicating that there was no significant difference between the measurements obtained via virtual reality when compared to those measured via a radiologist.

Conclusion: There was no significant difference between the indices measured via radiologic imaging, and those measured in virtual reality. This novel method provides guidance for how to develop and deploy digital methodologies for measurement in a digital 3D space. Similar to the benefit of improved mental representations of 3D models in VR, adding the ability to acquire measurements in 3D may allow for more relevant measurements unencumbered by 2D formats. Further research can be conducted to analyze whether this capability decreases operative time, or decreases complication rates.