Prospective in vitro accuracy and retrospective in vivo repeatability study.
Anglo-European College of Chiropractic, United Kingdom.
Continuous lumbar fluoroscopic sequences in vivo were obtained from 10 patients (mean age 38 years, SD 10.2, 42 % male) who had previously undergone OSMIA scans using a passive recumbent protocol which negated the action of muscle and motor control on inter-vertebral motion.
For accuracy, calibration model images of the lumbar spine with zero translation were acquired and others¡¯ imaged with digital fluoroscopy at 15 Hz during continuous flexion and extension. For inter-examiner repeatability, two observers blinded to each other's results analysed 10 in vivo lumbar spine fluoroscopy flexion and extension motion sequences for range of translation using automated tracking algorithms. For intra-examiner repeatability, one observer performed the analysis twice. Accuracy was calculated as the root-mean-square (RMS) difference between the known calibration model characteristics and the results acquired from the fluoroscopic sequences. Repeatability was calculated as agreement by standard errors of measurement (SEM) and 95 % limits of agreement and reliability by intra class correlation coefficients (ICC) for each inter-vertebral level (L2-S1) from the in vivo motion sequences.
The RMS error in measuring translation against the reference standard was under 0.8 mm in respect to a standard lumbar vertebra of 35 mm depth. With the exception of L5-S1 extension, the SEMs from the in vivo agreement studies were below 0.5 mm for all levels and directions (flexion-extension), and for reliability the ICCs were above 0.84.
OSMIA successfully measures inter-vertebral translation in vivo and in vitro in passive recumbent motion of the lumbar spine with greater repeatability and accuracy than previously reported studies using fluoroscopy.