Measurements agreement between low-cost and high-level handheld 3D scanners to scan the knee for designing a 3D printed knee brace

Use of additive manufacturing is growing rapidly in the orthotics field. This technology allows orthotics to be designed directly on digital scans of limbs. However, little information is available about scanners and 3D scans. The aim of this study is to look at the agreement between manual measurements, high-level and low-cost handheld 3D scanners. We took two manual measurements and three 3D scans with each scanner from 14 lower limbs. The lower limbs were divided into 17 sections of 30mm each from 180mm above the mid-patella to 300mm below. Time to record and to process the three 3D scans for scanners methods were compared with Student t-test while Bland-Altman plots were used to study agreement between circumferences of each section from the three methods. The record time was 97s shorter with high-level scanner than with the low-cost (p = .02) while the process time was nine times quicker with the low-cost scanner (p < .01). An overestimation of 2.5mm was found in high-level scanner compared to manual measurement, but with a better repeatability between measurements. The low-cost scanner tended to overestimate the circumferences from 0.1% to 1.5%, overestimation being greater for smaller circumferences. In conclusion, 3D scanners provide more information about the shape of the lower limb, but the reliability depends on the 3D scanner and the size of the scanned segment. Low-cost scanners could be useful for clinicians because of the simple and fast process, but attention should be focused on accuracy, which depends on the scanned body segment.