OBJECTIVES: This study assesses the ability to reconstruct costal cartilage images by using three-dimensional visualisation software (Mimics) based on semi-automated segmentation algorithm and to investigate its reliability and validity with an anthropometric analysis. DESIGN: Observational prospective study. SETTING: Plastic surgery department of a tertiary hospital. PARTICIPANTS: Twenty-two microtia patients who underwent autologous ear reconstruction. MAIN OUTCOME MEASURES: Preoperative thoracic computed tomography data were processed to Mimics software for three-dimensional costal cartilage imaging. The length, width, thickness and volume of the 9th costal cartilages were calculated from these images and compared with the direct measurements (DM) obtained intraoperatively. RESULTS: The intra-examiner reliability and inter-examiner reliability were high in terms of all four measurements (intraclass correlation coefficients, ICC: 0.876-0.984). There were no significant differences between image-based anthropometry and DM in the linear measurements except for the volume (P < .05). The mean volume calculation error of Mimics was -0.08 ± 0.13 mL. No correlation was found between the anthropometric variables and the absolute errors (P > .05). Furthermore, Bland-Altman plots were used to evaluate the agreement between the two methods. CONCLUSIONS: Despite a very small error was found in volume calculation, Mimics software was accurate and reliable in linear calculation. Three-dimensional costal cartilage imaging was found to be an efficient tool for morphological evaluation of costal cartilages. We believe that with the application of individualised cartilage models based on three-dimensional printing, the use of customised ear framework carving will be practicable in surgical training.
OBJECTIVES: This study assesses the ability to reconstruct costal cartilage images by using three-dimensional visualisation software (Mimics) based on semi-automated segmentation algorithm and to investigate its reliability and validity with an anthropometric analysis. DESIGN: Observational prospective study. SETTING: Plastic surgery department of a tertiary hospital. PARTICIPANTS: Twenty-two microtiapatients who underwent autologous ear reconstruction. MAIN OUTCOME MEASURES: Preoperative thoracic computed tomography data were processed to Mimics software for three-dimensional costal cartilage imaging. The length, width, thickness and volume of the 9th costal cartilages were calculated from these images and compared with the direct measurements (DM) obtained intraoperatively. RESULTS: The intra-examiner reliability and inter-examiner reliability were high in terms of all four measurements (intraclass correlation coefficients, ICC: 0.876-0.984). There were no significant differences between image-based anthropometry and DM in the linear measurements except for the volume (P < .05). The mean volume calculation error of Mimics was -0.08 ± 0.13 mL. No correlation was found between the anthropometric variables and the absolute errors (P > .05). Furthermore, Bland-Altman plots were used to evaluate the agreement between the two methods. CONCLUSIONS: Despite a very small error was found in volume calculation, Mimics software was accurate and reliable in linear calculation. Three-dimensional costal cartilage imaging was found to be an efficient tool for morphological evaluation of costal cartilages. We believe that with the application of individualised cartilage models based on three-dimensional printing, the use of customised ear framework carving will be practicable in surgical training.