Summary

3D curvature analysis can be used to visualize trochlea dysplasia (TP) and improve exisiting metrics quantifying TP.

Abstract

Introduction

Trochlea dysplasia significantly contributes to patellofemoral diseases, such as instability and osteoarthritis. Their treatment algorithms rely on an accurate description of the trochlea's morphology and groove. Current classification methods are primarily based on systems utilizing 2D images selected from MRI and CT, failing to capture the full extent of the three-dimensional (3D) trochlea. Anatomical 3D models created from segmented CT and MRI scans are an improvement compared to previous methods. Still, the interpretation of these models and derived metrics suffer from low interrater reliability.

Purpose

This study introduces a new method to visualize 3D models of distal femora using curvature analysis, allowing physicians to better understand the full complexity of trochlear dysplasia. The visualization is used both to qualitatively compare distal femora of patellofemoral instability (PFI) patients and controls and to significantly improve the reliability of the Entry Point to Trochlear Groove angle (EPTG) metric.

Study Design: Case-Control Study; Level of Evidence III.

Methods

3D models of 30 patients with recurrent patellar instability and 30 controls were created from CT-scans. Curvature was calculated to highlight the shape of proximal trochlear ridges and the trochlear groove by overlaying them on the 3D models. AP-views with and without curvature visualization were created, and used for qualitative comparison and to measure EPTG. EPTG was measured by two raters with and without the aid of the curvature maps. Significant differences between PFI patients and controls were compared with a Mann-Whitney U test. Interrater reliability was calculated using interclass correlation coefficients (ICC), classified according to literature and compared using a permutation test. Significance was assumed at 0.05. Results: Qualitative analysis between PFI patient knees (age: 23.9±8.4 years, female: 24, male: 6) and control knees (age: 21.8±5.6 years, female: 22, male: 8) showed that in general, PFI patients have a lateralized medial ridge and trochlear groove, with the trochlear groove being shorter and shallower. Qualitatively, differences between PFI patients and controls were significant for measurements both with and without the aid
of the curvature maps (p<0.001). Interrater reliability was significantly (p=0.0349) stronger when using the curvature visualization.

Conclusion

Curvature-based visualization aids overlayed on a 3D model have the power to increase the information gained from 3-D models, amplifying their potential value in clinical decision-making. Such visualizations facilitate both the identification of qualitative differences between patient and control morphology and the more reliable quantification of these differences through established metrics.