Abstract
Objectives:
The tibial tubercle to trochlear groove (TT-TG) distance often guides surgical decision- making for patellar instability. Traditional TT-TG measurements are limited by interrater reliability and lack of consideration for variable anatomy throughout flexion. The purpose of this study was to compare the traditional TT-TG measurement to three-dimensional measures generated from automated trochlear groove detection with sequential radial slices on magnetic resonance imaging (MRI).
Methods:
Twenty-four knees (4 healthy controls, 20 patellar instability patients) were segmented into three-dimensional distal femur and tibia models. A custom-designed program was employed to detect the trochlear groove through its course based on the shortest distance to the origin in a polar (radial) axis defined by the distal femoral condyles. The program also identified the surface projection of the tibial tubercle center of mass. The medial-lateral distance between the trochlear groove and tibial tubercle was calculated for the first 45 degrees of the detected groove to generate radial TT-TG (rTT-TG) values. The mean and maximum rTT-TG values were compared to the traditional TT-TG MRI measurement performed by one grader.
Results:
Mean rTT-TG values were significantly greater than traditional measurements (21.8 mm versus 17.9 mm, p=.003) with moderate correlation between measurements (Pearson r = 0.59, p=.002). In Dejour C and D dysplasia, the mean rTT-TG measurements trended towards being greater than traditional measurements (C: 24.3 mm versus 16.6 mm; D: 28.5 mm versus 21.8 mm; p=0.057 for both comparisons). Dejour C dysplasia demonstrated a significantly greater maximum rTT-TG than the traditional measurement (27.9 mm versus 16.6 mm, p=0.029), and Dejour D dysplasia demonstrated a similar trend (30.2 mm versus 21.8 mm; p=0.057).
Conclusions:
Traditional TT-TG distances may underestimate the true medial-lateral offset between trochlear groove and tibial tubercle, particularly in patients with high- grade dysplasia. This represents a potential alternative to the current standard measurement techniques which may improve measurement reliability by automation. Our study suggests that utilizing 3D radial TT-TG values via automated measurement may more accurately portray the variable anatomy of the trochlear entrance and offer a standardized alternative to a frequently used measurement to overcome reader reliability.
