Statistical shape modeling for digital nasal prosthesis design: Subjective and anthropometric validation

PURPOSE: Conventional nasal prosthesis design strongly depends on the artistic skills and experience of the anaplastologist. Digital approaches may reduce this dependency by automating patient-specific nasal shape generation. Statistical shape modeling (SSM) enables computation of a nose that aligns with an individual's facial morphology. MATERIALS AND METHODS: Fifty participants without facial deformities underwent three-dimensional (3D) facial scanning using an Artec Space Spider optical scanner. For each participant, the original nose (OG-nose) was digitally removed to create a standardized artificial nasal defect. An SSM-based algorithm reconstructed a nasal shape (SSM-nose), which was positioned on the facial model. A blinded assessment panel of participants, healthcare professionals, and patients evaluated randomized facial models featuring either the OG-nose or SSM-nose. Standardized questionnaires assessed natural appearance, aesthetics, notability, and facial harmony, and specific nasal characteristics (size, width, shape, depth, and pose). A paired-comparison questionnaire asked assessors to indicate their preferred nose. Objective comparison used predefined anthropometric measurements. RESULTS: Pooled analysis revealed significant differences for natural appearance and fit to the face, with SSM-noses rated lower than OG-noses. Case-wise comparisons, however, revealed no significant differences between OG-nose and SSM-nose ratings. In the paired preference assessment, healthcare professionals selected the OG-nose in 47% and the SSM-nose in 41% of cases; participants preferred the OG-nose in 71% and the SSM-nose in 17%. Several anthropometric parameters differed significantly, though absolute differences were small. Differences mainly concerned nasal projection, with reduced nasal bridge height (-1.9 mm), pronasale depth (-4.1 mm), and columella length (-3.9 mm), accompanied by corresponding angular changes, indicating slightly reduced protrusion in SSM-noses. CONCLUSIONS: SSM‑based nasal reconstructions were largely comparable to original noses, supporting the feasibility of this approach for digital anaplastology, although systematic under‑projection indicates the need for further model refinement.