Precision of an automated surface-based CT-RSA (asCTRSA) for assessing tibial implant migration: a porcine cadaver study and a clinical case

BACKGROUND AND PURPOSE: Radiostereometric analysis (RSA) is the reference standard for migration analysis, but adoption is limited by invasive marker implantation and specialized equipment. We developed an automated, surface-based CT-RSA (asCTRSA) method that reports a registration quality measure (mean rigid-body fitting error, ME). Our aim was to evaluate the precision by comparing asCTRSA with marker-based RSA and CT-based Micromotion Analysis (CTMA, v25.1). METHODS: Precision was assessed in a retrospective, repeated-measures, zero-migration study using a porcine cadaver tibial implant scanned in 7 positions (21 double examinations). Migration was reported as maximum total point motion (MTPM) for RSA and asCTRSA. CTMA provided maximum total translation (mTT) as an estimate of MTPM. Mean rigid-body fitting error (ME) was calculated for RSA and asCTRSA (acceptability threshold 0.35 mm). The minimal important difference for MTPM/mTT was 0.1 mm. asCTRSA was applied to 1 clinical case with known migration 3 and 8 months postoperatively (NCT04017533). RESULTS: In the porcine study, precision was 0.09 mm (95% confidence interval [CI] 0.02-0.18) for asCTRSA, 0.08 mm (CI 0.03-0.12) for CTMA, and 0.45 mm (CI 0.20-0.69) for RSA. In the clinical case, asCTRSA detected increasing migration (MTPM 2.5-3.6 mm) and exceeded CTMA's mTT by 0.4 mm at 8 months. Mean rigid-body fitting error remained below 0.35 mm. CONCLUSION: We showed that asCTRSA enables automated, surface-based CT-RSA and demonstrated higher precision than RSA in this zero-migration cadaver setting.