PURPOSE: To evaluate the accuracy of a smartphone application as a low-cost approach for digitizing a facial defect for 3D modeling. MATERIALS AND METHODS: A stone model of a facial defect was scanned using industrial computed tomography (reference scan) and was also scanned five times using a commercial laser scanner. A series of 24 sequenced digital photographs was taken five times by smartphone at two elevations. These images were uploaded and processed by a cloud-based server to create virtual 3D models. The 3D datasets were geometrically evaluated and compared to the reference data using 3D evaluation software. Mann-Whitney U test was used for statistical analysis, and the significance was set at P < .05. RESULTS: The overall mean 3D deviation ± standard deviation for the smartphone dataset was 604.9 ± 123.5 μm compared to 67.5 ± 0.49 μm for the laser scanner. There was a significant difference in the accuracy between the commercial laser scanner and the smartphone application (P = .009). CONCLUSION: The results showed that within the limits of this study and in reference to standard computed tomography imaging, data acquisition with a smartphone for 3D modeling is not as accurate as commercially available laser scanning.
PURPOSE: To evaluate the accuracy of a smartphone application as a low-cost approach for digitizing a facial defect for 3D modeling. MATERIALS AND METHODS: A stone model of a facial defect was scanned using industrial computed tomography (reference scan) and was also scanned five times using a commercial laser scanner. A series of 24 sequenced digital photographs was taken five times by smartphone at two elevations. These images were uploaded and processed by a cloud-based server to create virtual 3D models. The 3D datasets were geometrically evaluated and compared to the reference data using 3D evaluation software. Mann-Whitney U test was used for statistical analysis, and the significance was set at P < .05. RESULTS: The overall mean 3D deviation ± standard deviation for the smartphone dataset was 604.9 ± 123.5 μm compared to 67.5 ± 0.49 μm for the laser scanner. There was a significant difference in the accuracy between the commercial laser scanner and the smartphone application (P = .009). CONCLUSION: The results showed that within the limits of this study and in reference to standard computed tomography imaging, data acquisition with a smartphone for 3D modeling is not as accurate as commercially available laser scanning.