Pitfalls and Promise of 3-dimensional Image Comparison for Craniofacial Surgical Assessment

Three-dimensional (3D) photography is becoming widely used in plastic surgery. It provides an accurate and reproducible record of the facial surface anatomy and could be a versatile tool for treatment planning and assessment. However, the existing software tools available for the assessment of 3D facial imaging often give highly misleading results. The goal of this special topic article is to give clinicians an insight into methods of 3D image assessment and explain the reasons why results may be misleading. We point toward the advantages of an alternative approach using "nonrigid surface registration" for the comparison of pre- and postsurgical images. This approach is compared with the regular rigid surface registration, and this is illustrated by the assessment of a child with Crouzon syndrome before and after LeFort III osteotomy and distraction. Findings of the standard method imply that changes have occurred that are anatomically not possible, whereas the alternative approach indicates realistic changes. Furthermore, we demonstrate an exciting capacity of 3D image analysis to construct reference populations of normal head size and shape. These can be used to assess the parts of the head that are normal and abnormal pre- and posttreatment of the same child. We conclude that, while 3D image analysis has great potential in surgical assessment, existing software does not always give an adequate assessment. Collaboration among surgeons and engineering and computer science specialists should be encouraged. This way, more comprehensive and accurate techniques in patient assessment and surgical planning can be developed and applied in clinical practice.