Comparison of fusion assessment techniques: computer-assisted versus manual measurements

Abstract

Fusion assessment after cervical arthrodesis can be subjective. Measures such as bridging bone quantification or extent of (limited) motion on dynamic studies are common but difficult to interpret and fraught with biases. We compared manual measurement and computer-assisted techniques in assessing fusion after anterior cervical discectomy and fusion (ACDF). One hundred patients who underwent ACDF (512 intervertebral levels) were randomly selected for this radiographic review (follow-up 3-36 months). Two assessment techniques were performed by different observers, with each blinded to the results of the other. The manual spinous process displacement measurement technique was used to calculate motion between the spinous processes under magnification on a digital imaging workstation. Computer-assisted measurements of intervertebral angular motion were made using Quantitative Motion Analysis (QMA) software. Fusion criteria were arbitrarily set at 1 mm of motion for the manual technique and 1.5 degrees of angular motion for the QMA technique. The manual measurement technique revealed fusion in 61.7% (316 of 512) of the interspaces assessed, and QMA revealed fusion in 64.3% (329 of 512). These two assessment techniques agreed in 87.5% of cases, with a correlation coefficient of 0.68 between the two data sets. In cases in which the two techniques did not agree, QMA revealed fusion and the manual measurement revealed nonfusion in 64% of the disagreements; 98% of the disagreements occurred when motion was < 2 mm or 2 degrees. Although osseous fusion after arthrodesis remains difficult to assess, new computer-assisted techniques may remove the subjectivity generally associated with assessing fusion.