There may be notable differences in the trueness and precision of facial scans (FS) compared to virtual patient representations (VPR), according to a study recently published in the Journal of Prosthodontics.
Additionally, VPR accuracy may appear lower due to added errors in the alignment process and may not be reliable for more complex prostheses manufacturing purposes, the authors wrote.
"Data showed significant differences between the reference and virtual measurements for FS and VPR measurements," wrote the authors, led by Khaled Q. Al Hamad, MSc, of the Qatar University College of Dental Medicine (J Prosthodont, January 21, 2025).
In this in vivo study, a 27-year-old man underwent 30 VPRs using a custom intraoral scan body (ISB). Thirteen adhesive markers were added to the face as an extraoral scan body (ESB). Scans were performed with the ISB in place and without using an infrared laser sensor mounted on a tablet. Maxillary intraoral scans were also captured with and without the ISB, they wrote.
The scan files were imported into dental software, where facial and intraoral scans were aligned using the ISB and ESB as reference points. Measurements included 32 linear distances, 14 for FS accuracy and 18 for VPR accuracy, covering facial, lower face, and perioral regions. Trueness was assessed by the mean absolute difference (MAD) between virtual and direct measurements while precision was determined by standard deviation (SD).
Data were normally distributed except for 14 measurements. Most VPR measurements differed significantly from the reference (p ≤ 0.006), with overall trueness (MAD) of 2.76 and precision (SD) of 1.70. VPR measurements exceeded the reference in only seven instances, they wrote.
FS showed greater accuracy with trueness of 2.04, 1.66, and 0.80 and precision of 1.05, 0.92, and 0.91 for the total, lower face, and perioral regions, respectively. The VPR alignment workflow had lower trueness and precision than FS, with the lowest errors observed in the perioral region and higher errors associated with increasing measurement distance.
However, the study had limitations, including using only one subject. Future research should explore additional variables, such as diverse facial characteristics, gender differences, and skin tones, the authors added.
"VPR accuracy for the total and lower half of the face regions would be unreliable for the more demanding prostheses manufacturing," they concluded.
