Objective To investigate the diagnostic value of quantitative parameters of Synthetic magnetic resonance image compilation(Sy MRI)and least squares estimation and asymmetric echo iterative decomposition of water and fat imaging(IDEAL-IQ)sequence in the grade of rotator cuff injury.Methods 97 patients with clinical suspicion of rotator cuff injury and 28 healthy volunteers were prospectively collected,and general information(gender,age,BMI,duration of symptoms and visual analog pain score)was collected from all subjects.Conventional anatomical imaging sequences T1 WI FSE in axial view,T2 WI FS in axial view,T2 W FS in oblique coronal view,PDW in oblique sagittal view,magnetic resonance image compilation(Magic)sequence in oblique coronal view and IDEAL-IQ sequence in sagittal view were performed with MRI scanner.Based on the arthroscopic findings of the shoulder,the case groups were divided into 3 groups: tendinopathy group(grade II),partial tear group(grade III)and complete tear group(grade IV).The supraspinatus tendon was divided into lateral,medial and middle subregions according to its shape in oblique coronal T2 WI view,two radiologists measured the T1,T2 and PD values of the supraspinatus tendon,interclass correlation coefficient were used to compared the consistency between and within observers.Fat fraction(FF)of the supraspinatus,infraspinatus,subscapularis and teres minor was measured on Fat fraction images generated from the IDEAL-IQ sequence in the oblique sagittal position.One-way analysis of variance or Kruskal-Wallis H test were used to compare the differences of quantitative parameters in different regions,receiver operating characteristic(ROC)curves were drawn for parameters to evaluate the diagnostic efficacy for rotator cuff injury.The mean FF values of each muscle in different grades were compared by Friedman test with multiple related samples.The Spearman correlation analysis were used to analyze the correlations of quantitative parameters and the grade of rotator cuff injury.Results(1)There were 28,37,34 and 26 cases in the control group,tendinitis group,partial tear group and complete tear group,respectively.There was no significant difference in age,sex ratio,side and body mass index between the case group and the control group(P>0.05).(2)The ICC values of T1,T2 and PD values for the three subregions of the supraspinatus tendon were greater than 0.700.The differences in T1 values in the lateral subregion,T2 values in the lateral and middle subregions were statistically significant in the overall comparison across different classifications(all P < 0.001).The differences of T1 values in the middle and medial subregions,T2 values in the medial subregion and PD values in the lateral,middle and medial subregions were not statistically significant in the overall comparison of different grades of rotator cuff injury(all P>0.001).The quantitative parameters with statistically significant differences among grades were included in the ordered regression analysis,and the differences of T2 values in the lateral and middle subregions were statistically significant(all P < 0.001),and their regression coefficients were0.1160,0.1146,OR values(95% CI)were 1.123(1.037~1.216),0.122(1.151~1.197).The AUCs of the lateral subregions T2 values for diagnostic class I vs class IV,class II vs class IV and class III vs class IV were 0.891,0.797,0.723,95% CI was(0.801-0.981),(0.680-0.914)and(0.594-0.853)respectively,and the AUCs for the middle subregion T2 values diagnosing grade I vs IV,grade II vs IV,grade II vs III,and grade I vs III were 0.946,0.886,0.746,0.843,95% CI was(0.849-0.989),(0.809-0.962),(0.631-0.861),(0.745-0.941)respectively,all with high diagnostic efficacy.The T2 values in the lateral and middle subregions were positively correlated with the grading of rotator cuff injury,and the correlation coefficients were 0.542 and 0.615respectively(P<0.001).The T1 values in the three subregions,the T2 values in the medial subregions and the PD values in the three subregions were not significantly correlated with the grading of rotator cuff injury(P>0.001).(3)The FF values of supraspinatus,infraspinatus and subscapularis in the complete tear group were significantly higher than those in the normal group,tendinitis group and partial tear group,with statistically significant differences(P<0.05).There was no statistically significant difference among the FF values of supraspinatus,infraspinatus and subscapularis in the normal,tendinitis and partial tear groups(all P > 0.001).There was no statistically significant difference among the FF values of teres minor in the normal,tendinitis,partial tear group and complete tear group(Z=5.760,P=0.259).The supraspinatus FF values were positive correlated with rotator cuff injury grading(r=0.715,P <0.001),and the infraspinatus FF values were positive correlated with rotator cuff injury grading(r =0.628,P <0.001),the FF values of teres minor and subscapulari did not correlate significantly with rotator cuff injury grading(all P>0.001).There was no significant correlation between FF values of supraspinatus,infraspinatus,teres minor,subscapularis and VAS scores(all P >0.001).The FF values of supraspinatus were positive correlated with the duration of symptoms(r =0.710,P <0.001),and FF values of infraspinatus were positive correlated with symptom duration(r =0.587,P <0.001),and the teres minor FF values were not significantly correlated with symptom duration(r =0.306,P =0.004),and subscapularis FF values were positive correlated with symptom duration(r =0.379,P <0.001).Conclusions The T2 value of Synthetic MRI can be used as an effective quantitative parameter for the grading of rotator cuff injury,which can provide an objective basis for the clinical grading of rotator cuff injury.Fat fraction map of IDEAL-IQ sequence can achieve muscle fat quantification in rotator cuff injury,and fat infiltration in rotator cuff injury mainly involved supraspinatus,infraspinatus and subscapularis,and muscle fat infiltration was significantly higher in complete tears of tendons than in partial tears and tendinopathy. |