Accuracy evaluation of a lower-cost and four higher-cost laser scanners

• Published in: Journal of biomechanics Vol. 49; no. 1; pp. 127 - 131
• Main Authors: Campanelli, Valentina, Howell, Stephen M, Hull, Maury L
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 04.01.2016; Elsevier Limited
• Subjects: Accuracy; Algorithms; Animals; Bias; Blocking; Bones; Cattle; Computer Simulation; Conflicts of interest; Creaform; Diagnostic Imaging - economics; Diagnostic Imaging - methods; Error analysis; Femur - pathology; Gages; Imaging, Three-Dimensional; Laser Design; Laser scanner; Lasers; Models, Anatomic; Models, Statistical; Morphology; NextEngine; Nikon; Northern Digital; Orthopedics - methods; Physical Medicine and Rehabilitation; Repeatability; Reproducibility; Reproducibility of Results; Scanners; Scanning; Three-dimensional bone model; Repeatability; Three-dimensional bone model; Nikon; Northern Digital; Laser scanner; NextEngine; Creaform; Laser Design
• ISSN: 0021-9290; 1873-2380
• DOI: 10.1016/j.jbiomech.2015.11.015

Abstract Knowing the accuracy of laser scanners is imperative to select the best scanner to generate bone models. However, errors stated by manufacturers may not apply to bones. The three objectives of this study were to determine: 1) whether the overall error stated by the manufacturers of five laser scanners was different from the root mean squared error (RMSE) computed by scanning a gage block; 2) the repeatability of 3D models generated by the laser scanners when scanning a complex freeform surface such as a distal femur and whether this differed from the repeatability when scanning a gage block; 3) whether the errors for one lower-cost laser scanner are comparable to those of four higher-cost laser scanners. The RMSEs in scanning the gage block were 2 to 52 µm lower than the overall errors stated by the manufacturers. The repeatability in scanning the bovine femur 10 times was significantly worse than that in scanning the gage block 10 times. The precision of the lower-cost laser scanner was comparable to that of the higher-cost laser scanners, but the bias was an order of magnitude greater. The contributions of this study are that 1) the overall errors stated by the manufacturers are an upper bound when simple geometric objects like a gage block are scanned, 2) the repeatability is worse on average three times when scanning a complex freeform surface compared to scanning the gage block, and 3) the main difference between the lower-cost and the higher-cost laser scanners is the bias.