Related Experiment Video
Updated: Jun 12, 2025
Four-Dimensional CT Analysis Using Sequential 3D-3D Registration
Published on: November 23, 2019
Analyzing Alignment Error in Tibial Tuberosity-Trochlear Groove Distance in Clinical Scans Using 2D and 3D Methods
Johannes M Sieberer1,2, Nancy Park2, Albert L Rancu2
1Department of Mechanical Engineering, Yale School of Engineering and Applied Science, New Haven, Connecticut, USA.
Background:
Tibial tuberosity-trochlear groove distance (TT-TG) is often used as a primary metric for surgical decision-making in the treatment of patellofemoral instability (PFI), particularly when considering tibial tubercle transfer. Although TT-TG has high interrater reliability, it is prone to measurement differences caused by the alignment of the patient's leg in a scanner gantry, potentially influencing surgical decision-making. Quantification of this error within the clinical literature remains limited.
Purpose:
To quantify and specify the error in TT-TG caused by leg-scanner alignment by using detailed topographical landmarks and 3-dimensional (3D) analysis of computed tomography scans of patients with PFI.
Study Design:
Controlled laboratory study.
Methods:
Three-dimensional models of knees with PFI were created from computed tomography scans and used to identify TT-TG landmarks. TT-TG was measured using the established 2-dimensional (2D) and 3D methods. A model to estimate the differences between these 2 methods was created, and the orientation of the patients' legs in relation to scanner longitudinal axis was measured to validate this model via linear regression. Interrater reliability was calculated via intraclass correlation coefficients (ICC).
Results:
A total of 44 knees of patients with PFI were analyzed. Differences between the 2D and 3D methods ranged from -4.0 to 14.7 mm (mean ± SD, 2.7 ± 4.1 mm) with a root mean square difference of 4.8 mm. The TT-TG distance of the 2D method (19.8 ± 7.2 mm) was significantly (P = .045) longer than that of the 3D method (17.1 ± 4.9 mm). The variance of the 2D method was significantly larger than that of the 3D method. In total, 13 (29.5%) of the knees had a difference of >5 mm between 2D and 3D TT-TG. The estimation model had an adjusted r2 value of 1.00 and a resulting root mean square difference of 0.21 mm. 3D TT-TGs interrater reliability was good to excellent (ICC, 0.94 [95 CI%, 0.81-0.98]).
Conclusion:
3D TT-TG can be used to correct scanner-leg alignment errors, some of which are substantial when using only 2D TT-TG measurements.
Clinical Relevance:
The findings in this study suggest a need for caution and awareness of the potential effects of differences in alignment of the axes of the leg and scanner when using purely 2D TT-TG as a basis for surgical planning.
