RESEARCH REPORT PLATFORM PRESENTATION

DEVELOPMENT AND VALIDATION OF THE CLARITYSMART SPINAL MOTION ANALYSIS RESEARCH TECHNOLOGY. Abbott J¹, McCane B²; ¹New Zealand Centre for Physiotherapy Research, University of Otago, PO Box 56, Dunedin 9054, New Zealand. ²Department of Computer Science, University of Otago, PO Box 56, Dunedin 9054, New Zealand

PURPOSE: To develop an accurate and time-efficient computer-assisted method for measuring displacement kinematics of lumbar spine segments from flexion-extension radiographs. RELEVANCE: Lumbar segmental mobility disorders, such as lumbar segmental instability and lumbar segmental rigidity, are believed to be associated with recurrent and chronic low back pain. Research of these disorders requires accurate measurement of lumbar segmental kinematics to diagnose the presence of a disorder, or measure change in movement characteristics. PARTICIPANTS: A subset of 138 patients with recurrent or chronic low back pain, 30 asymptomatic volunteers, and dried human vertebral specimens. METHODS: Development of the computer assisted measurement methodology started with a literature review identifying previously developed methodology. We critiqued the published methods and critical discussions, incorporating the most successful strategies into our developing plans. We developed a Java-based computer program for estimating the planar transformation of a rigid body. Version 3 of our program used the iterative closest point algorithm to perform least-squares matching of cubic spline curves which were placed in a graphic user interface by a mouse or pen digitiser. The parameters computed were angle of rotation, finite centre of rotation (FCR), centre of reaction, and translation. We compared our methodology to the methods of Bogduk et al, Frobin et al, and Crisco et al. ANALYSIS: We tested the reliability and validity of our methodology and calculations in five phases: concurrent validity of rotation was tested against the NIH Image program; concurrent validity of translation was tested against manual constructions provided by an independent researcher; validity of rotation and finite centre of rotation were tested by computer-simulated trials, and finally by an in vitro error study; reliability was assessed by analysing the results of repeated measures. RESULTS: Intraclass correlation coefficient (ICC) values for agreement of rotation measurement between ClaritySMART and NIH Image were ICC3,4 0.98 (95% confidence interval 0.92, 1.00). Agreement of centre of reaction and translation measurements between our method and the method of Bogduk et al. were ICC3,1 0.99 (0.97, 1.00) and 0.98 (0.97, 1.00) respectively. Validity assessed by computer-simulated trials demonstrated our new method to result in significantly lower angle error (p < 0.001) and FCR error (p < 0.05) than competing methodologies. In vitro true FCR error was significantly lower than competing methodologies (p < 0.001, and p < 0.02) at less than 7.3mm (sd 8.3). Intra-rater reliability coefficients (ICC3,1) were: rotation 0.97 (0.95, 0.98); translation 0.89 (0.80, 0.94); it is necessary for mathematical reasons to exclude cases with rotation <1.0° to achieve satisfactory reliability for FCR measurements of 0.82 (0.67, 0.91) for the x axis, and 0.66 (0.41, 0.81) for the y axis, and centre of reaction 0.75 (0.57, 0.87). CONCLUSIONS: Our new methodology for measurement of displacement kinematics of lumbar segmental motion is more accurate and more efficient than previously available methods. We recommend it for research involving lateral flexion-extension radiographs, and have made it freely available at www.claritysmart.com. IMPLICATIONS: This new methodology significantly reduces the error of measuring motion from flexion-extension radiographs, making possible research designs that were previously not possible due to limitations of older technology. KEYWORDS: low back pain, lumbar spine, segmental motion, measurement, displacement, kinematics. FUNDING ACKNOWLEDGEMENTS: University of Otago Research Grants. CONTACT: haxby.abbott@otago.ac.nz

ETHICS COMMITTEE: Otago and Canterbury Regional Ethics Committees; University of Otago Ethics Committee