From the Division of Plastic and Reconstructive Surgery, Department of Surgery, Yale School of Medicine.
Plast Reconstr Surg. 2021 Apr 1;147(4):915-924. doi: 10.1097/PRS.0000000000007776.
The "accuracy" of virtual surgical planning across multiple procedure types is not known. The authors aimed to compare the planned outcome from virtual surgical planning to the actual postoperative outcome for five craniofacial procedure types performed by a single surgeon: implant cranioplasty, cranial vault remodeling, orthognathic surgery, mandible reconstruction, and mandibular distraction.
Stereolithography formats were obtained from virtual surgical planning and compared to postoperative computed tomographic scans for consecutive patients who underwent one of the five procedure types. Volumetric renderings of the operated bony region of interest were overlaid and compared using a Boolean operation to compute conformity (as a percentage of the region of interest). Conformity across procedure type was analyzed using analysis of variance and post hoc Bonferroni analysis, where appropriate.
One hundred thirty patients were included (51.5 percent male and 49.5 percent female; mean age, 27 years; 59 orthognathic surgery, 32 cranial vault remodeling, 16 mandible reconstruction, 12 mandibular distraction, and 11 implant cranioplasty patients). The highest tier of conformity was obtained for implant cranioplasty (median, 76.8 ± 10.3 percent) and mandible reconstruction (mean, 69.4 ± 11.2 percent), followed by orthognathic surgery (mean, 55.0 ± 7.3 percent) and mandibular distraction (median, 41.9 ± 20.3 percent), followed by cranial vault remodeling (mean, 22.2 ± 12.1 percent) (p < 0.001 between tiers and p > 0.05 among tiers).
Virtual surgical planning resulting in custom permanent implants and intraoperative guides provides more predictable results compared to virtual surgical planning used for procedures involving higher degrees of skeletal repositioning and postoperative movement (i.e., mandibular distraction and nonrigid cranial vault remodeling). In cases with expectedly lower conformity, excellent outcomes can be achieved with sound intraoperative judgment.
目前尚不清楚多种手术类型的虚拟手术规划的“准确性”。作者旨在比较同一位外科医生进行的五种颅面手术类型（植入物颅骨成形术、颅盖重塑、正颌手术、下颌骨重建和下颌骨牵引）的虚拟手术规划和实际术后结果。
从虚拟手术规划中获得立体光刻格式，并将其与接受五种手术类型之一的连续患者的术后计算机断层扫描进行比较。对感兴趣的手术骨区域进行体积渲染，并使用布尔运算进行叠加和比较，以计算一致性（以感兴趣区域的百分比表示）。使用方差分析和适当的事后 Bonferroni 分析分析跨手术类型的一致性。
共纳入 130 例患者（男性占 51.5%，女性占 49.5%；平均年龄 27 岁；59 例正颌手术、32 例颅盖重塑、16 例下颌骨重建、12 例下颌骨牵引和 11 例植入物颅骨成形术患者）。一致性最高的是植入物颅骨成形术（中位数为 76.8%±10.3%）和下颌骨重建（平均值为 69.4%±11.2%），其次是正颌手术（平均值为 55.0%±7.3%）和下颌骨牵引（中位数为 41.9%±20.3%），最后是颅盖重塑（平均值为 22.2%±12.1%）（不同级别之间的差异有统计学意义（p＜0.001），不同级别之间的差异无统计学意义（p＞0.05））。
与涉及更高程度的骨骼重新定位和术后运动（即下颌骨牵引和非刚性颅盖重塑）的手术相比，虚拟手术规划产生定制的永久性植入物和术中导板可提供更可预测的结果。在预计一致性较低的情况下，通过良好的术中判断仍可获得出色的结果。
