# The Templating Software Spectrum: Unveiling Variations in the 3D Assessment of Glenoid Morphology in Shoulder Arthroplasty Planning

## TL;DR

This study compares how different 3D planning software affects measurements of shoulder bone shape, showing that results can vary significantly between platforms.

## Contribution

The study reveals that commonly used 3D planning systems for shoulder surgery produce inconsistent measurements, particularly for inclination and retroversion.

## Key findings

- MS showed greater dispersion and weak correlations for retroversion compared to other systems.
- Inclination measurements were precise but showed platform-specific differences, with ZS having notably lower values.
- Inter-system correlations for inclination were strong among BP, EE, and ZS but moderate for MS.

## Abstract

Introduction

Three-dimensional (3D) computed tomography planning improves preoperative quantification of glenoid morphology when planning total shoulder arthroplasty, but measurement reliability across commonly used planning platforms remains uncertain.

Methods

A retrospective within-subject comparison of 30 shoulders was planned on four systems: Stryker Blueprint (BP), Zimmer Signature (ZS), Exactech Equinoxe (EE), and SurgiCase Materialise (MS). Measurements of retroversion and inclination were analysed to evaluate measurement variability and correlations between systems.

Results

Retroversion medians (IQR) were BP −6.5 (10.75), ZS −6.0 (7.0), EE −5.5 (7.0), and MS −3.5 (19.5); overall p=0.37. MS showed greater dispersion and weak correlations with other systems (ρ≈0.03-0.07). Inclination means (SD) were BP 5.93 (8.13), ZS 1.28 (7.27), EE 3.53 (7.57), MS 6.00 (5.07); overall p<0.001. All paired contrasts were significant except for BP vs. MS (−0.07, 95% CI −2.14 to 2.01; p=0.948). Inter-system inclination correlations were strong among BP, EE, and ZS (r≈0.80-0.90) and moderate to strong for MS vs. others (r≈0.57-0.75).

Conclusion

Platforms were not fully interchangeable. Retroversion showed no group-level bias but reduced precision, particularly for MS, while inclination was highly precise yet displayed platform-specific level differences (notably lower with ZS). Based on this, threshold-adjacent decisions (augment selection, baseplate tilt) may vary by platform. Verification of retroversion outliers and confirmation of inclination near boundaries are recommended.