Related Experiment Video
Updated: Jun 19, 2026
Pedicle Screw Placement Using an Augmented Reality Head-Mounted Display in a Porcine Model
Published on: May 24, 2024
Does applying CPAK classification boundaries using measurements from a surgical navigation system result in different
Matthew D Hickey1, Asim Khan2, Alexander Tham2
1School of Biomedical Engineering The University of British Columbia Vancouver British Columbia Canada.
Purpose:
Coronal Plane Alignment of the Knee (CPAK) classification defines knee phenotypes using two factors: arithmetic hip-knee-ankle angle (aHKA) and joint line obliquity (JLO) measured via long leg radiographs (LLRs), which can produce errors due to variability in patient positioning. Navigation-based technology has been shown to increase accuracy and may be used where LLRs are not routine. Given this, we asked: (1) Does applying the CPAK classification system to measurements obtained from a surgical navigation system result in different class distributions when compared to LLRs? and (2) Given a mechanical alignment approach, what is the accuracy of a navigation-based system in delivering the targeted postoperative CPAK Type V classification?
Methods:
We analysed data from 4087 total knee arthroplasty (TKA) procedures undertaken using surgical navigation at our institution between March 2007 and October 2022. Of these patients, 80 had preoperative LLRs from which medial proximal tibial ankle (MPTA) and lateral distal femoral angle (LDFA) were measured to calculate aHKA and JLO. We compared the mean preoperative and postoperative distributions of these alignment variables and classified patients into CPAK types and compared CPAK class frequencies between pre and postoperative TKA.
Results:
The mean signed differences between preoperative MPTA and LDFA values measured using surgical navigation and LLRs were 0.2° (σ = 2.8°, p = 0.46) and 1.4° (σ = 2.6°, p < 0.01), respectively. The mean signed differences in preoperative aHKA and JLO were calculated to be -1.7° (σ = 3.7°, p < 0.01) and 1.2° (σ = 3.8°, p < 0.01). Postoperatively, 87.3% of cases were classified as Type V.
Conclusions:
While surgical navigation yields population-level CPAK distributions similar to LLRs, individual class assignment was preserved in only 38.8% of patients showing that measurement modality frequently alters phenotype classification. Discrepancies likely arise from differing angle definitions, cartilage assessment techniques and the non-weight-bearing nature of navigation compared to standing radiographs.
Level Of Evidence:
Level III.
