The treatment for DDH is to achieve concentric reduction of femoral head and acetabulum, so as to restore the normal anatomical structure of hips, which is also the key to evaluate the therapeutic effect after surgeries [17, 18]. Most scholars believe that the FNA increase is one of the important pathological changes of DDH [19]. Proximal femoral derotation osteotomy is often used in surgeries for children with DDH to correct the large FNA, so as to achieve concentric reduction of femoral head and acetabulum [20, 21]. However, to determine whether or not a rotation is needed or what the rotation angle is, accurate measurement is required before surgeries. The traditional methods mostly depend on the surgeon's experience to determine the osteotomy rotation angle, which is obviously lack of accuracy and will affect the surgery effect [22]. Therefore, finding an accurate and simple method to measure FNA before surgeries is essential.
In the past, many scholars have studied how to accurately measure FNA and invented many FNA measurement methods. However, it is a pity that so far there is no gold standard recognized by most people [23]. At present, CT measurement methods are widely used. There are many different kinds of CT measurement methods [24, 25], where 2D-CT measurement method is the most widely used because of its relatively lower technical requirements, but there is a great controversy about the accuracy of the measurement results. A lot of authors think that 2D-CT method is to use 2D images to measure the 3D structure, which is obviously inaccurate, as 2D images cannot simultaneously show the complete structure of the proximal femur [26]. Besides, in 2D-CT measurement method, it is more random and subjective in selection of CT slice images for measurement, which leads to great differences in measurement results between different physicians [27, 28]. In this study, it is found that the difference in results between 3D-PF model and 2D-CT measurement was statistically significant, indicating that the accuracy of 2D-CT measurement method was not very well. In the 2D-CT method for the same 68 patients, the results obtained by the three observers were quite different. In addition, in the 2D-CT measurement for the same patient by the same observer, the difference were still statistically significant in results between the first measurement and the second performed 3 months later, indicating that the stability and repeatability of 2D-CT measurement method in FNA measurement were poor.
We further analyzed the 2D-CT measurement processes by the three observers, and found that for the same patient, the CT slice images selected by the three observers for measurement were not exactly the same. In addition, in the marking for femoral neck axis and line connecting medial and lateral femoral condyles, the mark points selected are also different between observers. Therefore, we believe that the instability, poor accuracy and poor repeatability in FNA measurement results of 2D-CT method may be caused by the difference in selection of CT slice images and mark points between the observers.
With 3D-CT technology, we can reconstruct 2D images into 3D images by imaging software. By rotating the 3D images, we can observe the whole from any angle, and correct the patient position according to the skeletal coordinates before measurement. The 3D-CT FNA measurement method used in this study is our self-designed method based on the characteristics of the femurs [12]. This method is easy to operate and easy to learn. Since applying this method in the clinic, we have achieved good therapeutic effect, because we calculated the rotation angle of proximal femoral osteotomy based on the FNA measured by 3D-CT method before surgeries, and performed osteotomy according to the calculation result mentioned above. Previous studies have shown that the stability and repeatability in measurement results of our 3D-CT method are significantly better than that of 2D-CT method, but some scholars still question the accuracy of our measurement results [13]. It is difficult to accurately measure the femur structure in human body, and DDH patients are children, so we cannot take the real femur for accurate FNA measurement. Therefore, we have been unable to get the accurate FNA to test the accuracy of 3D-CT measurement method.
The rapid development of 3D printing technology provides an idea for us to test the accuracy of 3D-CT method in measuring FNA in children with DDH. We collected the CT data of femurs of the 68 children with DDH, and then printed a 1:1 femur model exactly the same as the human skeleton by 3D printing technology, and then performed physical measurement on the model to obtain the accurate FNA of DDH patients. In this study, the FNA in children with DDH measured by 3D-CT method was compared with the real FNA obtained by 3D-PF method, and it was found that there was no statistical difference in the measurement results between 3D-CT method and 3D-PF method, indicating that the FNA in children with DDH measured by 3D-CT method was very accurate.
Whether the FNA on the dislocated sides of DDH patients increases is controversial [29]. By measuring FNA of the 68 patients with 3D-PF method, we found that the average FNA of dislocated hips was 48.5 ± 6.6°, and that of normal hips was 36.9 ± 13.1°. There was significant difference between the two results as revealed by statistical test, indicating that the FNA on the dislocated hips in children with DDH generally increases, and the FNA on the dislocated hips in 3–8 years old children with DDH is 11.6° larger than the normal hips. Besides, after assigned by age, we found that compare the 3 years old group with the 8 years old group, the FNA of the normal hips had no significant difference. However, the FNA of the dislocated hips enlarged from 42.7 ± 9.9°to 51.6 ± 8.2° significantly.