Bi-planar calibration method for templating of hip joint arthroplasty: phantom study and proof of concept.

Purpose: Calibration of radiographs is a critical step in digital templating for hip arthroplasty. Calibration errors of > 1.5% lead to over- or undersizing of the templated implants and may affect logistics and patient safety. Contemporary calibration methods are known to be imprecise with average errors of 6.5% and wide variance. A novel bi-planar radiograph-based calibration method is proposed, and a phantom study was conducted as proof of concept.

Methods: A spherical external calibration marker (ECM) is placed in front of the pubic symphysis of a pelvic bone model at twelve different positions. For each marker position, standard anteroposterior radiographs and four corresponding lateral radiographs with different degrees of rotation (0°-30°) are taken (overall, 60 radiographs). Calibration factors are calculated for an internal calibration marker (ICM) at the centre of the right hip (reference) and the ECM using a novel algorithm. Rotation and marker positions simulate foreseeable use errors and misplacements and aim to test robustness of the method against these errors.

Results: ECM calibration factor was 125.9% (range 124.7-127.2), and the mean ICM calibration factor was 126.6% (range 126.2-127.1) ([Formula: see text]). Four images (8.3%) were beyond the 1% error threshold (all with 30° rotation). The mean difference was 0.79% (SD 0.49).

Conclusion: The bi-planar method precisely predicts the true calibration factor of the hip joint plane under various conditions. In lateral radiographs, rotation of up to 20° did not adversely affect the precision and all images had calibration errors below the threshold for clinical significance.