Validation of a simple computerized tool for measuring spinal and pelvic parameters.

OBJECT

Recent studies have emphasized measuring the sagittal vertical axis (SVA) and pelvic parameters (pelvic incidence, sacral slope, and pelvic tilt) when evaluating spinal disorders. An accurate and reproducible measurement is important for a reliable result. Although computerized measurement is more consistent than manual measurement, computerized measurement requires an expensive software program, the need to transfer images to a workstation, and additional education for users. An inexpensive and convenient computerized measurement program is desirable and necessary. The object of this study was to propose a computerized tool for measuring spinal and pelvic parameters and to evaluate the efficacy of this new tool compared with manual measurement.

METHODS

The authors devised a tool that provides computerized measurements of the SVA and pelvic parameters in a picture archiving and communication system (PACS) without transferring images to another program. This tool was created by merging functions in the PACS. The resulting tool is easy to implement by merging functions (indicate the center of 2 points, plot a vertical and a horizontal line from a point, and measure the angles between lines) in any image viewer. The tool was made into icons on a toolbar in the PACS. Measurements of distance and angle were computerized by identifying crucial points after selecting the icon. For SVA, 4 points were identified around each corner of the C-7 body and a fifth point at the superior/posterior corner of the S-1 body. For pelvic parameters, 4 points were identified at the centers of each femoral head and at the anterior/superior and posterior/superior corners of S-1. Thirty-three whole-spine lateral radiographs were randomly selected from the radiographic database. To evaluate inter- and intraobserver variability between observers and method, skilled (2 years of experience) and unskilled (1 week of experience) observers measured SVA and pelvic parameters 3 times with a 7-day interval between each time using both computerized and manual measurement methods. The reliability was measured using the intraclass correlation coefficient.

RESULTS

The computerized method showed better congruity than the manual method in both skilled and unskilled observers (p < 0.05), and the intraclass correlation coefficients were > 0.9. The skilled observer showed better agreement than the unskilled observer with both computerized and manual methods, and this difference was prominent in measuring pelvic parameters (p < 0.05). The computerized method required less time than the manual method, especially for the unskilled observer (p < 0.05).

CONCLUSIONS

A computerized measurement of pelvic parameters may be a more reliable and efficacious approach than manual measurements. This benefit is more prominent in the unskilled observer, and adding this simple function to an image viewer may be recommended in future studies.