| 1BackgroundNon-traumatic osteonecrosis of the femoral head (NONFH) is a debilitating disease that commonly affects young adults in their third to fifth decade of life. It is also the most common reason for total hip replacement in many regions especially in Asian countries.The prognosis of osteonecrosis is highly related to the extent and location of the lesion involving the femoral head. When the lesions involve more than30%of the femoral head or in the weight-bearing zone, most of the hips will progress to collapsing if left untreated. However, if the extent of the lesions is small, many of them will remain asymptomatic and some of them may resolve as demonstrated by magnetic resonance (MR) image analysis. It is therefore important to estimate the extent of involvement accurately in order to identify those who need to be treated and those who need to be observed. When collapse of necrotic bone occurs, reduced hip joint function and pain cause a significant impairment of activities of daily living. Although hip replacement surgery is an effective means for improvement of pain and joint function, long term outcomes, particularly in patients at the prime years of life, are not yet determined. Joint preservation surgery such as core decompression or femoral osteotomy may be another effective method of therapy; however, its outcome largely depends on the size and location of the lesion, and satisfactory outcomes cannot be achieved at an advanced stage of collapse. Therefore, early and reliable assessment of the size and location of necrotic lesions before progression to collapse and the prognosis for collapse is important in ONFH.There are plenty of measurements of lesion based on X-ray film and MRI in clinical,including the arc of involvement, necrotic area ratio and necrotic volume ratio.It is not conformed which measurement is the most effcctive and accurate one. It is also rarely reported of comparison and optimization of measurements.Magnetic resonance (MR) imaging is widely used in the diagnosis of ONFH for its high sensitivity and specificity for the disease. Crucial factors affecting the natural course and therapeutic results are the size and site of the necrotic region in the femoral head and several authors have tried to estimate the extent of necrosis quantitatively using MR imaging. T1-weighted spin-echo (SE) imaging was used exclusively in these reports.Thus the same study based on T1-weighted spin-echo (SE) imaging was performed by us. In the study by Plakseychuk et al, the reliability and reproducibility of the current staging systems were demonstrated to be poor and with high intraobserver and interobserver variations. The variations mostly come from the different results of necrosis volume calculation by different methods. Possibly, different authors acquire different results in the same condition. Therefore,no accurate measurement accepted has been for clinical application so far. The difference of the measured results may also be induced by the type of sample type, the type of therapeutic interventions and other factors.2ObjectiveThe purpose of this investigation is to master the distribution of lesions in the coronal magnetic resonance (MR) images,and to optimizate a more practical and effective measurement,by measuring and analysising the necrosis area of the early-middle stage non-traumatic osteonecrosis of the femoral head(NONFH) based on magnetic resonance (MR) images.At the same time, this investigation may provide the basis for intervention plus TCM administration by the associativity between femoral head necrosis area and therapeutic effect.3Method3.1According to the inclusion standard and elimination standard,we screened the cases of NONFH,managed by intervention plus TCM administration in Wangjing hospital of China Acadamy of Chinese Medicine Science from January2011to January2012,with the final diagnosis ONFH ARCO Ⅱ and follow-up time over12months. There were87patients(120hips), including59men(78hips) and28women(42hips). The age is from21to68years,and the averge age is40.59±10.56years.3.2All the MR images were taken by the digital camera and saved in the format of JPEG. The original method used digital processing of the images with computer software. In this study, the MR images were imported into the AutoCAD2007software with magnification adjustment. In the coronal T1-weighted MR images,the level where the femoral neck was exposed completely was set as the middle one,marked as level0(L0).In the same way, the first before the middle was L1, the second L2, the third L3and the fourth L4; and the first after the middle was L-1, the second L-2,the third L-3,the forth L-4.3.3The necrotic lesions were distinguished by the specialists in the visual tool of the AutoCAD2007software in a Microsoft Windows XP operating system. 3.4The necrotic lesions, defined as the region with abnormal signals, were outlined by the tracing tool of the AutoCAD2007software in a Microsoft Windows XP operating system. We had a calculation of osteonecrotic area using two-dimensional projections and made comparison between the current volumetric measurement methods.3.5Necrotic volume ratio method:Necrotic volume was acquired by the sum of necrotic area multiplied by thickness;Femoral head volume was acquired by the sum of segment area multiplied by thickness.Then necrotic volume ratio was acquired by the division between necrotic volume and femoral head volume.Necrotic area ratio methodrNecro area and segment area were outlined by the tracing tool of the AutoCAD2007software; necrotic area ratio was acquired by the division between necro area and segment area.Mean necrotic area ratio method:According to the distribution of lesions in the coronal magnetic resonance (MR) images, the levels of more lesions were included, excluding those of less lesions one by one.Then mean necrotic area ratio was acquired by the average of necro area ratios in all the levels included.7levels method was L0、1、2、3、4、-1、-2included.6levels method was L0、1、2、3、4、-1、included.51evels method was L0、1、2、3、-lincluded.4levels method was L0、1、2、3included.3levels method was L0、1、2or L1、2、3included.3.6We collected patients’clinical data、Harris Hip Score(HHS) and imaging assessment data, set up Databases, then analysised data,and conform the correlationgs between mean necrotic area ratio method and effect of intervention plus TCM administration.4Results4.1In the coronal MR images, lesions mainly distributed in four levels:109(90.83%)in the first layer before the middle;109(90.83%)in the second layer before the middle;101(84.17%) in the middle layer;99(82.50%) in the third layer before the middle.4.2In three-dimensional measurements, necrotic volume ratio calculated by MRI-based segmentation method was0.333±0.151in87cases(120hips).4.3In two-dimensional measurement methods, mean necrotic area ratio method of the middle, the first, second, third and fourth layer before the middle and the first, second layer after the middle was0.312±0.147; mean necrotic area ratio method of the middle, the first, second, third and fourth layer before the middle and the first layer after the middle was0.333±0.153; mean necrotic area ratio method of the middle, the first, second and third layer before the middle and the first layer after the middle was0.348±0.172; mean necrotic area ratio method of the middle, the first, second, and third layer before the middle was0.365±0.174; mean necrotic area ratio method of the middle, the first, second layer before the middle was0.377±0.202; mean necrotic area ratio method of the first, second and third layer before the middle was0.348±0.172.4.4Through ANOVA test, there is no statistical significance (P>0.05) in difference between mean necrotic area ratio method of the middle, the first, second, third and fourth layer before the middle and the first, second layer after the middle and MRI-based segmentation method.Through ANOVA test, there is no statistical significance (P>0.05) in difference between mean necrotic area ratio method of the middle, the first, second, third and fourth layer before the middle and the first layer after the middle and MRI-based segmentation method.Through ANOVA test, there is no statistical significance (P>0.05) in difference between mean necrotic area ratio method of the middle, the first, second and third layer before the middle and the first layer after the middle and MRI-based segmentation method.Through ANOVA test, there is no statistical significance (P>0.05) in difference between mean necrotic area ratio method of the middle, the first, second and third layer before the middle and MRI-based segmentation method.Through ANOVA test, there is statistical significance (P＜0.05) in difference between mean necrotic area ratio method of the middle, the first and second or the first、second、third layer before the middle and MRI-based segmentation method.4.5Among the120hips,12were graded as small-sized lesions (<15%of involvement),43were graded as mediumsized lesions (15%-30%of involvement), and65were graded as large-sized lesions (>30%of involvement), by using the MRI segmentation method. With the use of mean necrotic area ratio method of the middle, the first, second and third layer before the middle,10were graded as small-sized lesions (<15%of involvement),32were graded as mediumsized lesions (15%-30% of involvement), and78were graded as large-sized lesions (>30%of involvement). Through Chi-square test, there is no statistical significance (X2=2.617, P=0.106) in difference between mean necrotic area ratio method of the middle, the first, second and third layer before the middle and MRI-based segmentation method.The overall agreement on the grading was high.4.6According to the Harris titre standard,Grade A:HHS increased by11.62±7.46before and after treatment; Grrade B:HHS increased by11.29±1.84before and after treatment; Grade C:HHS increased by9.63±9.11before and after treatment. The increase in Grade C is lowest.According to image titre standard, Grade A:the image stabilization rates was100%after treatment; Grrade B:the image stabilization rates was84.41%after treatment; Grade C:the image stabilization rates was79.53%after treatment. Grade C is lowest.5Conclusion5.1Necrotic lesions most were in L1、L2、L0、L3.5.2Femoral head necrotic area ratio method is simple effective and relatively accurate.5.3There were correlations between the effect managed by intervention plus TCM administration and necrotic area ratio. |
PDF Full Text Request