Treatment methods in patients with scoliosis depend on the Cobb angle in coronal plane and morphology of the sagittal planes
[11, 13–16]. Therefore, a veracious measurement is pivotal for the options of treatment. The Cobb angle measurement in the coronal plane has been studied fully, and the accuracy and reliability were good. Owing to variable measurement criteria, manual measurement errors, and difficulty in visualizing measurement landmarks in the measurement of the spina-pelvic alignment, the accuracy and reliability is often difficult and poor as previous studies demonstrated
[9, 10]. Therefore, developing a reliable method of radiographic measurement of the sagittal–pelvic alignment other than the traditional manual method is indispensable.
Computer-based SurgimapSpine measurement technique can open plain radiographs photographed or scanned, which form the digitized image. It adjusts image contrast and brightness enabling a better identification of key anatomical parameters not normally available for measurement on traditional radiographs. SurgimapSpine measurement technique has some advantages such as the following: rapid comparison between radiographs taken at different times of a patient, cheap storage, and images formatted by photos not films. In the present study, we found no significant difference in the intraobserver and interobservers’ reliability between the manual and the SurgimapSpine methods in the coronal Cobb’s angle measurement. However, the intra/interobserver reliability of the sagittal alignment found in SurgimapSpine tool was significantly better than those in manual method, especially in T2–T5, PI, and PT. Our data showed that the reliability of both the intraobserver and interobservers match well with the SurgimapSpine method and is more reliable in the Cobb angle measurement in the sagittal plane. When SurgimapSpine software is used for Cobb angle measurement, important parts of the spine can be enlarged and seen more clearly by changing the contrast, and the borders of the vertebrae can be enhanced by computerized options; after drawing lines through the endplates of end vertebrae, the software measures the angle automatically, which may reduce sources of error. Therefore, Cobb angle measurement by SurgimapSpine software both in coronal and sagittal alignment may be more accurate when compared with those measured with the manual method.
The coronal Cobb angle is usually used for the assessment and treatment of scoliosis. The excellent overall reliability of Cobb angle measurement has been well studied
[1–3, 8, 10, 17]. For the undefined end vertebra setting, Gstoettner
 found a mean ICC of 0.97 for the intraobserver and interobserver reliability measurement by the manual method, whereas for the computer-assisted method, a mean ICC value for interobserver and intraobserver reliability was 0.93 and 0.96, respectively. Although measurement of Cobb’s angle using computer-assisted method was slightly better than that of the manual method, the computer-assisted method does not improve the measurement accuracy. Our study found similar excellent levels of intraobserver and interobserver reliability for the Cobb measurement by both manual and SurgimapSpine methods. These data suggested that the use of SurgimapSpine measurements does not improve measurement accuracy of the Cobb angle. In the end vertebrae defined setting, ICC of coronal Cobb angle in the current study was comparable to previous results
[8, 18] in which the end vertebrae were undefined, suggesting that the end vertebral selection was not an important factor in reliability of the Cobb measurement. Although different end vertebrae may result in Cobb angle variability, they do not influence accuracy of measurement. Therefore, in clinical practice it is not necessary to ensure the same end vertebrae.
In a study on sagittal–pelvic measurement for 29 normal young adults, John et al.
 found the intraobserver ICC obtained by manual measures for PI was 0.69, PT 0.60, SS 0.77, and LL 0.90, and the ICC for interobserver was 0.41, 0.42, 0.64, and 0.57, respectively. Pearson correlation coefficient between computer-aided measures and manual measures for PI was 0.59, PT 0.63, SS 0.72, and LL 0.68. The authors concluded the reliability of computer-aided measures was notably higher in all parameters except for LL where the difference between the manual measures and the computer-aided measures was not obviously different. In the present study on AIS (adolescent idiopathic scoliosis) radiography using manual measurement, the reproducibility and reliability for T2–T5 and PT was only fair to good, whereas those angles of T5–T12, T10–L2, LL, PI, and SS were measured with excellent reproducibility both in intraobserver and interobserver. However, with regard to the intraobserver reproducibility and interobserver reliability of the SurgimapSpine tools, all parameters measured were excellent. Intra/interobserver reliability/reproducibility for T2–T5 thoracic kyphosis was markedly worse than for all other measures either in manual measures or in SurgimapSpine tool. Another study focusing on reliability of manual measures in AIS patients found intraobserver (0.22–0.83) and interobserver for T2–T5 (0.33–0.47) reliability was generally poor. However, other sagittal parameters were excellent
. The reliability and reproducibility of T2–T5 and PT Cobb angle measurement using both manual tool and SurgimapSpine tool in our study were disappointing. Other sagittal radiographic measures demonstrated good to excellent correlation. Causes for poor reliability of T2–T5 and PT Cobb angle may be related to the overlying density of the upper thoracic rib cage and scapula and femoral head. Our data also showed that sagittal measurement with SurgimapSpine tool obviously increases reproducibility and reliability, especially in Cobb angle of T2–T5 and PT. However, T2–T5 measurement is still not satisfactory.
Although the differences in the two methods are very small, the results of the present study indicate that the SurgimapSpine measurement is an equivalent measuring tool to the traditional manual in coronal Cobb angle, but is markedly advantageous in spino–pelvic measurement especially in T2–T5 and PT.