ABSTRACT
BACKGROUND: Penile phenotype in hypospadias is currently assessed visually or manually (e.g., ruler, goniometer) for clinical, education, and research applications. However, these methods lack precision and accuracy across raters and cannot be reevaluated retrospectively following a surgical repair. The project aim was to evaluate the precision and reliability of penile dimensions obtained from digital and three dimensional (3D) printed models created from intraoperative (OR) structured light scans (SLS) during primary pediatric penile procedures. METHODS: Boys ages 1 month to 6 years underwent first- or single-stage penile surgery at a single institution were enrolled in this prospective study (IRB #20-000143). For each patient, immediately following placement of a stay suture under consistent manual tension, intra-operative dimension measurements with a ruler were obtained. A digital 3D model was created prior to penile repositioning using an Artec Space Spider scanner and Artec Studio 13 software. Following the case, two different raters completed 10 digital measurements of each generated model in Autodesk Fusion 360. These digital models were subsequently 3D printed and two different raters completed 10 manual dimension measurements of each 3D printed model using a ruler. A one-way random effects intraclass correlation coefficient (ICC) evaluated measures of agreement between and within raters, respectively. Analyses were conducted in R version 4.2. RESULTS: Six scans were obtained (hypospadias: 4, circumcision: 2). Intra-rater assessments showed excellent precision across repeated digital measurements; manual measurements of 3D printed models had excellent reliability for glans width and penile length but poor to good reliability for glans height. Inter-rater reliability was good to excellent for glans width (0.77-0.95) and good for penile length (0.71-0.88). However, there was poor inter-rater reliability for glans height (0-0.14). Following training regarding glans height location, there was an improvement in precision and repeatability of manual and digital measurements. CONCLUSION: Digital measurement of OR-derived 3D models resulted in excellent repeatability for each rater and improved between-rater reliability over manual measurement of 3D printed models alone, ensuring that images can be compared by various surgeons both now and in the future. SLS is promising as a novel modality to digitally generate 3D models, thereby informing phenotypic analysis for research and education. Further development of digital measurement methods to ensure consistency between raters for quantitative assessment of additional parameters and assessment of the technology within the pre-operative environment for surgical planning is planned.