Image-Based Quantitative Osteotomy Planning for Congenital Proximal Radioulnar Synostosis: Plan-Execution Discrepancy and Postoperative Radiographic Outcomes.

Abstract

Proximal radioulnar synostosis is a rare congenital anomaly that restricts forearm rotation and significantly impairs daily function. Surgical correction typically involves radial osteotomy with shortening and derotation; however, planning of osteotomy level, bone shortening, and rotational correction remains largely subjective and relies heavily on surgeon experience. This study retrospectively evaluated an image-based quantitative planning framework designed to provide objective geometric parameters for osteotomy in congenital proximal radioulnar synostosis. Preoperative anteroposterior and lateral radiographs from eight unilateral cases were reconstructed to derive patient-specific deformity geometry, central bone axes, cut points, predicted cutting angles, and model-estimated radial shortening. The planning algorithm uses polynomial fitting and calculus-based minimization to identify osteotomy points. These predictions were compared with intraoperative measurements and postoperative radiographic and functional outcomes. All procedures were complication-free, and seven of eight forearms showed improved rotational range. At final follow-up, two forearms had concentric radial head alignment, whereas six showed mild subluxation or dislocation. Observationally, larger discrepancies between model-estimated and surgeon-applied shortening were associated with suboptimal postoperative radial head alignment. In one representative case, a substantial discrepancy corresponded with progressive radial head dislocation despite preserved clinical motion. This study highlights the potential value of quantitative, image-based modeling to support osteotomy planning by providing transparent, reproducible parameters that may reduce variability in surgical decision-making. Although limited by its small sample size, retrospective design, and reliance on radiographs, this work establishes a foundation for future prospective validation of computational decision-support tools in pediatric orthopedic deformity correction. Level IV.