Multi-Modality MR Images Of Sacroiliac Arthritis In Early Diagnosis And Active Phase Evaluation

Ankylosing spondylitis(AS)is an autoimmune disease that usually begins in the sacroiliac joint and culminates in calcification and ossification of the axial skeleton,leading to the deformity of joints in AS,which severely affects the patient’s quality of life.Early inflammation in AS starts from the sacroiliac joint,and the main manifestations are low back pain,morning stiffness,peripheral joint pain and other nonspecific clinical manifestations.Developments in imaging and genetics could enable the early detection and early treatment of inflammation.The detection of early-stage AS is needed for clinical treatment.X-ray and/or computed tomography(CT)examinations are insufficiently sensitive to detect early-stage AS:the results are often negative(nr-axSpA),and the ionizing radiation involved makes these examinations unsuitable for patient follow-up review.Magnetic resonance imaging(MRI)does not involve ionizing radiation,can realize high-resolution imaging of soft tissue,and is feasible for multi-parametric and multi-directional imaging.In the guidelines of the Assessment of Spondyloarthritis International Society(ASAS),MRI is recommended for sacroiliac joints in the early stage of AS.Despite a high positive detection rate of bone marrow edema(BME),chronic bone structural lesions could be easily missed and misdiagnosed.In recent years,using texture analysis(TA)has exhibited superior performance for medical image analysis due to powerful image feature recognition capabilities.Deep learning has been reported to offer outstanding advantages for identifying BME signs of sacroiliitis,but little research has been performed to the diagnosis of early sacroiliitis.Therefore,in the first part of this study,oblique axial T1-weighted(T1WI)and oblique fluid-sensitive Fat-saturated(Fs)T2-weighted(T2WI)MR images were used to explore the application value of TA to diagnosising early sacroiliitis.Ankylosing spondylitis(AS)has individual heterogeneity.Some patients present with only mild symptoms,while others have severe damage and a poor prognosis,with a rapidly progressive disease course.The presence of frequent attacks during the active phase is often indicative of disease progression and requires early and active treatment.In contrast to the widely accepted clinical guidelines for monitoring the activity of rheumatoid and systemic lupus erythematosus during the active phase,assessing AS disease activity is particularly difficult due to the lack of objective clinical or biochemical markers that can be identified and quantified.In current clinical practice,AS activity is primarily scored using the Bath Ankylosing Spondylitis Disease Activity Index(BASDAI),which relies on subjective,self-assessed patient reports but can be quickly obtained through patient self-assessment.The Working Group for The Assessment of Spondyloarthritis International Society(ASAS)recommended combining the Ankylosing Spondylitis Disease Activity Score(ASDAS)scoring system and MRI to be more objectively and accurately.Therefore,the second part of this paper analyzes the value of dynamic contrast-enhanced(DCE)-MRI and ZOOMit-Diffusion weighted imaging(DWI)for the assessment of sacroiliitis activity.Quantitative parameters include:ZOOMit-DWI apparent diffusion coefficient(ADC),DCE volume transfer constant(Ktrans),rate constants(Kep),extravascular volume fraction(Ve).Semi-quantitative parameters include:initial area under the time concentration curve(iAUC),factor of enhancement(Fenh),Slope of enhancement(Senh),Time to peak(TTP).Part I Differntiation of Early Sacroiliitis Based on TIWI and FsT2WI Texture AnalysisPurpose:By analyzing the MR T1WI and FsT2WI image texture features,we aim to evaluate the value of texture analysis models in the early diagnosis of sacroiliac arthritis.Methods:Patients who underwent MR examination of sacroiliac joints in our hospital from January 2015 to January 2018 were retrospectively selected and enrolled.The modified New York Criteria for AS were used.Patients were classified into the nr-axSpA group if the digital radiography(DR)and/or CT results within 7 days from the MR examination showed a DR and/or CT grade<2 for bilateral sacroiliac joints,or a DR and/or CT grade<3 for a unilateral sacroiliac joint.Patients were classified into the r-axSpA group if the DR and/or CT grade was 2 to 3 for bilateral sacroiliac joints or the DR and/or CT grade was 3 for a unilateral sacroiliac joint.Patients were considered to have a confirmed diagnosis if the DR or CT grade was 4 for sacroiliac joints and were thereby excluded.A control group of healthy individuals matched in age and sex to the patients was included in the study.First,two radiologists independently qualitatively scored oblique coronal T1WI and FsT2WI nonenhanced sacroiliac joint images.The diagnostic efficacies of the two radiologists were judged and compared using an assigned Likert score,conducting Kappa consistency test of diagnostic results between two readers.Further drew Receiver operating characteristic curve(ROC)and calculated the area under the curve(AUC).Texture analysis models(T1WI-TA model and FsT2WI-TA model)were constructed through feature extraction and feature screening.The Delong test was used to compare the diagnostic efficacy between both TA models and between the TA models and the readers.Results:A total of 32 patients in nr-axSpA group,30 in r-axSpA,group and 30 in healthy control group were included in the study.The qualitative scores of the two radiologists could significantly distinguish between healthy controls and nr-axSpA groups,nr-axSpA and r-axSpA groups(both P<0.05).There was no significant difference in the differential diagnoses of the two radiologists between the healthy control and nr-axSpA groups(AUC:0.7880 vs 0.8050,P=0.838)and between the nr-axSpA and r-axSpA groups(AUC:0.7490 vs 0.7110,P=0.645).Both TA models could significantly distinguish between healthy controls and nr-axSpA group,nr-axSpA group and r-axSpA group(both P<0.05).There was no significant difference in the differential diagnoses of the two TA models between the healthy control and nr-axSpA groups(AUC:0.934 vs 0.976,P=0.1 838)and between the nr-axSpA and r-axSpA groups(AUC:0.917 vs 0.848,P=0.2592).In terms of distinguishing between the healthy control and nr-axSpA groups,both the TA models were superior to the qualitative scores of the two radiologists(all P<0.05).In terms of distinguishing between the nr-axSpA and r-axSpA groups,the T1 WI-TA model was superior to the qualitative scores of the two radiologists(P=0.023 and P=0.007),whereas there was no significant difference between the FsT2WI-TA model and the qualitative scores of the two radiologists(P=0.134 and P=0.065).Conclusion:Based on MR imaging,the T1WI-TA and FsT2WI-TA models were highly effective for the early diagnosis of sacroiliac joint arthritis.The T1WI-TA model significantly improved sacroiliac arthritis early diagnostic efficacy compared to the qualitative score of radiologists,while the FsT2WI-TA model was comparable to that of radiologists.Part Ⅱ The value of DCE-MRI and ZOOMit-DWI for the assessment of sacroiliitis activityPurpose:To investigate the value of quantitative and semiquantitative DCE-MRI parameters and ZOOMit-DWI apparent diffusion coefficient(ADC)values in the differential diagnosis of active and inactive sacroiliitis in patients with confirmed sacroiliitis.Methods:Patients with confirmed sacroiliitis based on ASAS guidelines from April 2017 to December 2018 were included and grouped according to the ASDAS-CRP score,with ASDAS-CRP≥2.1 corresponding to the active group and ASDAS-CRP<2.1 corresponding to the inactive group.Sex-and age-matched healthy individuals were recruited as healthy controls.Independent predictors of sacroiliitis activity were screened based on quantitative DCE-MRI parameters(Ktrans,Kep,Ve),semiquantitative parameters(iAUC,Fenh,Senh,TTP),and ZOOMit-DWI ADC values in the bone marrow region and synovial region in patients and controls.The diagnostic efficacy of the independent predictors was determined by a receiver operating characteristic(ROC)curve.Results:A total of 71 patients with axial spondyloarthritis(axSpA)were included,and 26 healthy individuals were included in the control group.Quantitative parameters:sacral(Ktrans,Ve,ADC),iliac(Ktrans,Ve),and synovial(Ktrans)parameters were all higher in the active group than in the inactive group(all P<0.05);Semiquantitative parameters:sacral(iAUC,Fenh),iliac(iAUC,Fenh),and synovial(iAUC)parameters were all higher in the active group than in the inactive group(all P<0.05);Multivariate logistic regression analysis confirmed that sacral Ktrans,sacral Fenh,iliac iAUC and synovial iAUC were independent predictors for differentiating active versus inactive phases,with areas under the ROC curve of 0.9317,0.7937,0.8794 and 0.7124 respectively.Conclusion:Quantitative and semiquantitative DCE-MRI parameters(sacral Ktrans,sacral Fenh,iliac iAUC and synovial iAUC)are independent predictors of the sacroiliac joint activity phase,so DCE-MRI is expected to provide a non-invasive and accurate basis for the activity prediction of sacroiliitis.