Dependable Automated Approach for Measuring the Retrograde Superior Ramus Screw Corridor in Pelvic Fracture Fixation.

Abstract

Background: Precise measurement of the intraosseous corridor within the superior pubic ramus is essential for the accurate percutaneous placement of a retrograde superior ramus screw (SRS). However, conventional manual measurement methods are often subjective, leading to variations in results among observers. Our goal was to develop an automated and dependable method for determining the retrograde SRS corridor.

Methods: We developed an automated technique that utilized a computed tomography (CT) image-based search algorithm to identify the retrograde SRS corridor with the maximum diameter. We evaluated the reliability of this automated approach in comparison to a manual method using 17 pelves. Subsequently, we used both methods to measure the diameter, length, and orientation of the retrograde SRS corridor in 204 pelves in a Chinese population and assessed the intra- and interobserver agreement of each method by calculating the root-mean-square error (RMSE) and constructing Bland-Altman plots. We determined the screw applicability (percentages of hemipelves that could be treated with specific sizes of screws) for each method. Additionally, we investigated potential factors influencing the corridor, such as sex, age, height, and weight, through regression analysis.

Results: The intra- and interobserver intraclass correlation coefficients (ICCs) for the automated method (0.998 and 0.995) were higher than those for the manual approach (0.925 and 0.918) in the assessment of the corridor diameter. Furthermore, the diameter identified by the automated method was notably larger than the diameter measured with the manual method, with a mean difference and RMSE of 0.9 mm and 1.1 mm, respectively. The automated method revealed a significantly smaller corridor diameter in females than in males (an average of 7.5 and 10.4 mm, respectively). Moreover, use of the automated method allowed 80.6% of the females to be managed with a 4.5-mm screw while a 6.5-mm screw could be utilized in 19.4%, surpassing the capabilities of the manual method. Female sex had the most substantial impact on corridor diameter (β = -0.583).

Conclusions: The automated method exhibited better reliability than the manual method in measuring the retrograde SRS corridor, and showed a larger corridor diameter for screw placement. Females had a significantly smaller corridor diameter than males. Given the intricate nature of the automated approach, which entails utilizing different software and interactive procedures, our current method is not readily applicable for traumatologists. We are working on developing integrated software with the goal of providing a more user-friendly solution for traumatologists in the near future.

Level of evidence: Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.