Objective To evaluate the imaging quality and diagnostic value of different MR sequences for abnormalities of knee cartilage. Methods Three sequences of MR, including double echo steady state (DESS), multi echo data imagine combination (MEDIC) and true fast imaging with steady state precession (TrueFISP) were performed in 18 healthy volunteers. Signal-to-noise ratio (SNR) of knee cartilage and contrast-to-noise ratio (CNR) to surrounding tissues were measured and compared. Results SNR of cartilage was the highest in MEDIC sequence imaging, CNR of cartilage-synovia was the highest in DESS sequence imaging among three sequences (P<0.05). Conclusion Compared with the other sequences, DESS seems to be optimal in diagnosis of abnormalities in knee cartilage.

Objective To observe MRI features of the localized intra-articular form of pigmented villonodular synovitis (PVNS) within the knee joint. Methods Fifteen patients with pathologically proved localized intra-articular form of PVNS were retrospectively reviewed, and the MRI appearances were observed. Results Solitary nodule or mass within the knee joint was detected in all 15 patients. The lesion located in the infra-patellar fat pads in 11 patients. Most lesions were round or oval, with a well-defined margin. The maximum diameters ranged from 2.0 cm to 6.0 cm, with a mean value of 3.1 cm. The signal intensity of these lesions was characteristically heterogeneous on SE T1WI, TSE PDWI, and TSE T2WI, and the whole signal was similar to that of muscles on SE T1WI and TSE PDWI, but a little higher on TSE T2WI. Prominent low signal intensity on T2WI because of deposition of hemosiderin was found within 14 lesions, and 9 of 15 lesions showed characteristically low signal marginal rim. Conclusion MRI features of the localized intra-articular form of PVNS include a well-defined solitary nodule or mass locating in the infra-patellar fat pad, heterogeneous signal intensity, and prominent interior focal low signal or peripheral low signal rim on T2WI.

Objective To compare the diagnostic value of conventional MR imaging with MR arthrography for acetabular labral tears at 3.0 T MR imaging Methods Forty-four patients suspected of acetabular labral pathology were prospectively scanned with conventional MR imaging and MR arthrography of the affected hip joints. All MR data were retrospectively studied. Each acetabular labrum was divided into three regions of anterior, superior, and posterior, and the acetabular tear and its classification were recorded. Five patients underwent subsequent hip arthroscopy. Wilcoxon signed ranks test was used to compare the difference between MR imaging and MR arthrography, and Kappa value was adopted for the evaluation of agreement of these two techniques. Results One hundred and sixteen regions of acetabular labrum were evaluated to be concordant at MR imaging and MR arthrography. Only 16 regions had different results at MR imaging and MR arthrography, in which 9 were diagnosed to have tears at MR arthrography but non-tear at MR imaging, and 6 had tears at MR imaging but non-tear at MR arthrography, and the other one had tears at both exams but of different classifications. In 5 patients confirmed by hip arthroscopy, both the evaluations of MR imaging and MR arthrography were accordant with that of arthroscopy. There were no statistical difference between MR imaging and MR arthrography ( Z = 0. 347, P ＞ 0. 05 ) for diagnosing the acetabular labral tears, and good agreements (Kappa = 0. 781 ,P ＜ 0. 01 ) were found between these two methods. Conclusion Conventional MR imaging of hip joints at 3.0 T shows similar diagnostic value with MR arthrography in evaluating the acetabular labral tears.

Objective To evaluate the diagnostic value of MR arthrography for rotator cuff tear. Methods A prospective study was undertaken in 32 patients with shoulder diseases using MR arthrography. Abnormalities revealed on MR arthrography of the rotator cuff, the labrum, and the tendon of the long head of the biceps brachii muscle were recorded. The results of MR arthrography were compared with those of arthroscopy and surgery. Results Based on the results of arthroscopy and surgery, 32 patients comprised 14 complete rotator cuff tear, 6 partial rotator cuff tear, and 12 subjects without tear. For detecting rotator cuff tear, the sensitivity, specificity, and accuracy of MR arthrography were 100%, 100%, 100% respectively. For detecting complete rotator cuff tear, the sensitivity, specificity, and accuracy of MR arthrography were 100%, 94%, 97% respectively. Meanwhile, MR arthrography revealed all multi tendon tear and all abnormalities of the labrum and the tendon of the long head of the biceps brachii muscle before arthroscopy. Conclusion MR arthrography is an accurate and comprehensive method for evaluating rotator cuff tear.

Objective To determine the val ue of five MR imaging signs in diagnosing the bucket handle tears (BHT) of menisci of the knee. Methods MR imaging of 139 knees with subsequent arthroscope exams were retrospectively evaluated. Based on the results of arthroscopy, 19 knees had BHT of menisci. Two radiologists evaluated each MR exam independently, with discrepancies resolved by consensus. Each MR exam was analyzed for five signs: a double posterior cruciate ligament sign, a flipped meniscus sign, an absent bow tie sign, an internal displaced fragment sign, and an abnormal circumferential meniscus sign. Sensitivity, specificity, positive and negative predictive values, and accuracy for diagnosing BHT of menisci were calculated for the presence of each individual sign. Results The sensitivities of these five signs ranged between 52.6% and 89.5%, and specificities ranged between 83.3% and 98.3%, respectively. Positive and negative predictive values ranged between 42.9% and 88.2%, and between 92 7% and 98.3%, respectively. The accuracy of these five signs ranged between 82.7% and 96 4%. Conclusion BHT of menisci may have many signs on MR imaging. The diagnostic sensitivities of the internal displaced fragment sign and the abnormal circumferential meniscus sign are the highest. The double posterior cruciate ligament sign has the highest specificity and positive predictive value, while the internal displaced fragment sign has the highest negative predictive value and accuracy.

Objective To determine the diagnostic value of indirect MRI arthrography for meniscal tears.Methods 75 cases of routine MRI of the knee and 64 cases of indirect MRI arthrography of the knee were retrospectively reviewed.These patients all had subsequently arthroscopic examination.Based on the results of arthroscopic examination,the diagnostic value for meniscal tear of routine MRI and indirect MRI arthrography were compared.Results For medial meniscal tears,the indirect MRI arthrography gave a sensitivity of 72%,specificity of 100%,accuracy of 89.1%,positive predictive value of 100% and negative predictive value of 84.8%.For lateral meniscal tears,the indirect MRI arthrography gave a sensitivity of 83.8%,specificity of 90.9%,accuracy of 87.5%,positive predictive value of 89.6% and negative predictive value of 85.7%.No matter for medical meniscal tears or lateral meniscal tears,the diagnostic differences between the indirect MRI arthrography and the routine MRI were not statistical significant.Conclusion Although the diagnostic value of indirect MRI arthrography for meniscal tears was quite high,it was not different with routine MRI statistically.

Objective To summarize MRI features of anterior cruciate ligament graft tear and to explore the differences of MRI findings between acute tear and chronic tear, and compare the diagnostic ability of MRI and clinical examinations for graft tear. Methods MR images of 43 patients (44 knees) with anterior cruciate ligament graft tear(40 complete tear, 4 partial tear)confirmed by secondary arthroscopy were retrospectively analyzed. There were 18 acute tear and 26 chronic tear. Primary and secondary signs reported with conventional anterior cruciate ligament tear were adopted to evaluate graft tear. The exact probability method was used to compare the prevalence difference between various direct and indirect signs and the χ2 test was used to compare the accuracy between MRI and physical examination. Results The primary signs in MR images of anterior cruciate ligament graft tear included graft discontinuity in 13 kness, graft thickening with edematous high signal intensity in 12 knees, decreased slope of graft fibers in 6 kness, graft disappearing in 5 knees, and distinct graft atrophy in 3 knees. The secondary signs included kissing bone contusion in 4 knees, posteriorcruciate ligament buckling in 3 knees, increased anterior tibial displacement in 2 knees, bone contusion of the lateral condyle of femur, and bone contusion of thetibia condyle in 1 knee, respectively. There were no significant differences regarding the proportion of each sign between acute and chronic graft tear. Accuracy of MRI, Lachman test, and anterior drawer test were 87.5%(35/40), 95.0%(38/40)and 95.0%(38/40), respectively, which were all significantly higher than that of pivot shift test(42.5%,17/40) with significant differences(χ2=17.80, P<0.0083). Conclusions MRI is sensitive for diagnosing anterior cruciate ligament graft tear, the primary signs is the main evidence for the diagnosis of ACL graft tear, but it is hard to distinguish acute and chronic graft tear based on MR findings. The diagnostic accuracy has no statistically significant differences among MRI, Lachman test, and anterior drawer test, but they are all higher than pivot shift test.

Objective To investigate the value of gadobenate dimeglumine (Gd-BOPTA) enhanced MRI in evaluation of reserved liver function.Methods Totally 68 patients were included among 143 patients who underwent Gd-BOPTA en hanced MRI.The enhancement ratios of SNR and CNR of right lobe,left lobe and caudal lobe were calculated on the ima ges from liver acquisition with volume acceleration flex sequence (LAVA-Flex) of hepatobiliary phase.The average values of right and left lobes and of three lobes were calculated and compared.According to Child-TurcottePugh (CTP) classification,the patients were divided into Class A and Class B.Integrated model of end-stage liver disease (iMELD) was applied to divide the patients into iMELD≤30 and iMELD＞30.The differences of SNR and CNR enhanced ratios were com pared,and the relationship between enhancement ratios and iMELD scores was analyzed.Results There were no significant differences in the enhancement ratios of SNR and CNR among each lobe,the average of two or three lobes (all P＞0.05).The SNR and CNR enhancement ratios of Class A were higher than those of Class B (both P＜0.05).The SNR and CNR enhancement ratios of iMELD≤≤30 were higher than those of iMELD＞30 (both P＜0.05).The SNR and CNR enhancement ratios showed negative correlations with iMELD scores (SNR:r=0.29,P-0.02;CNR:r=-0.32,P=0.01).Conclusion The intracellular uptake of Gd-BOPTA decreases with impaired liver function.Measurement of the de gree of parenchymal enhancement of any lobe or the whole liver on the images of Gd-BOPTA enhancement MRI in the hepatobiliary phase might reflect the reserved liver function.

Purpose To investigate the feasibility of T1 measurement using multi-slice inversion recovery turbo spin echo (IR-TSE) and variable flip angle three-dimensional fast low angle shot (3D FLASH) for full-joint coverage delayed gadolinium enhanced MRI of the cartilage. Materials and Methods MR phantoms using dilated gadolinium solution with different concentrations and ex vivo bovine cartilages were scanned using multi-slice IR-TSE and variable flip angle 3D FLASH sequences and the T1 values were calculated. The referring T1 value were obtained using single slice IR-TSE sequence. Results On phantom study, the correlation coefficiencies between multi-slice IR-TSE, 3D FLASH sequences and single slice IR-TSE was 1.000 and 0.997, respectively, where P<0.001 on ex vivo bovine cartilage study, the T1 values of the trypsin degraded species measured by single-slice, multi-slice IR-TSE and 3D-FLASH were significantly lower than the control species. The correlation coefficiency between single-slice and multi-slice IR-FSE on the control group, the trypsin degraded group and both was 0.821(P=0.012), 0.968(P=0.001), and 0.953(P=0.001) respectively. The correlation coefficiency between single-slice IR-FSE and 3D-FLASH on the trypsin degraded group, control group and both was 0.199(P=0.637), 0.757(P=0.030), and 0.775(P=0.001) respectively, where P<0.001. Conclusion Both Multi-slice IR-TSE and variable flip angle 3D-FLASH sequences can be applied for T1 measurement of the full joint cartilage.

Objective To determine the value of MRI for diagnosing the posterior meniscus root tear. Methods MR examinations of 30 patients with tear of the posterior meniscus root confirmed by knee arthroscopies were retrospectively reviewed. Of the 30 patients, 17 with posterior medial meniscus root tear (MMRT) and 13 with posterior lateral meniscus root tear (LMRT). The diagnostic sensitivity of MRI for the posterior meniscus root tear was analyzed. Fisher's exact test was used to compare the detection rate of MRI for MMRT with that for LMRT. Results All 17 cases with MMRT and 9 cases out of 13 with LMRT were correctly diagnosed by MRI and the diagnostic sensitivity of MRI for the posterior meniscus root tear was 86. 7% (26/30). The main MR appearance of the posterior meniscus root tear was distortion of the meniscal root, with its low signal replaced by abnormal high signal. The detection rate of MRI for MMRT (17/17)was significantly greater than that for LMRT (9/13) ( P =0. 026). The prevalence of MMRT associated with meniscus extrusion ( 15/17 ) was significantly greater than that of LM RT ( 6/13 ) ( P = 0. 020), but the prevalence of MMRT associated with anterior cruciate ligament (ACL) injury (5/17) was significantly lower than that of LMRT (11/13) (P = 0. 004). Conclusion MRI is a relatively good method for detection of posterior meniscus root tears and associated injuries.