【作者】 李锦青；

【导师】 张绍祥；

【作者基本信息】 第三军医大学 ， 计算医学， 2012， 博士

【摘要】 研究背景冻结肩(frozen shoulder，FS),又称“粘连性关节囊炎”(adhesive capsulitis,AC)，根据美国肩肘外科医师学会定义：冻结肩是一种特定的肩关节囊疾病，是一类引起盂肱关节僵硬的粘连性关节囊炎，表现为肩关节周围疼痛，肩关节各个方向主动和被动活动度降低、受限，影像学检查除骨量减少外无明显异常的疾患。在中国等亚洲地区称为“肩周炎”或“五十肩”，好发于40-60岁中老年，女性较男性多见。目前国外文献中以“冻结肩”、“粘连性关节囊炎”两种定义为多，并逐渐被广泛接受。目前，冻结肩主要根据病史和体格检查做诊断，“冻结肩是个临床诊断”。由于其发病机制、病理及病因未知，使得目前对冻结肩的诊断及治疗仍存在很大的盲目性。而作为相对金标准的肩关节镜诊断冻结肩却是侵入性检查，所以无创的影像学检查诊断冻结肩变得至关重要，人们开始寻找冻结肩的客观影像学证据。国外一些学者对冻结肩影像学诊断做了有益的初步探索：X线肩关节造影肩关节腔容量小于10ml，腋隐窝消失时，可考虑冻结肩诊断；超声检查发现患者喙肱韧带(coracohumeral ligament, CHL)增厚，肩袖间隙（rotator Interval，RI）低回声背景及血管增多、血供增加，可考虑冻结肩诊断。磁共振逐步成为冻结肩主要的检查方法，但报道结果不一致、没有得到互认：有研究认为腋囊厚度超过4mm可以作为诊断冻结肩的一个重要标准；也有研究认为腋囊厚度大于3mm(斜冠状位不压脂T2WI)提示冻结肩；有研究认为冻结肩关节囊滑膜血供增加，RI出现强化的纤维血管瘢痕组织、肱二头肌长头腱(long head of the bicepstendon,LBT)起点区软组织增厚、腋囊增厚是诊断冻结肩重要的征象；有研究认为喙肱韧带增厚、肩袖间隙关节囊增厚、喙突下脂肪三角完全闭塞是诊断冻结肩的重要征像；有研究认为喙肱韧带增厚与肩关节外旋、内旋受限有关。既往研究表明肩关节囊、肩袖间隙及其内容物CHL已成为冻结肩磁共振研究者关注的靶部位。既往磁共振文献探讨了喙肱韧带增厚对冻结肩的诊断作用，但是，CHL厚度测量依赖于CHL显示率，而CHL显示率又依赖于肩袖间隙内高信号脂肪对比，因此，研究肩袖间隙脂肪分布情况，对于冻结肩磁共振诊断至关重要，但是罕有正常志愿者肩袖间隙脂肪分布、喙肱韧带显示率的报道，缺乏肩袖间隙脂肪分布、喙肱韧带显示率与冻结肩发病相关性的研究。既往磁共振文献探讨了肩袖间隙造影检查异常强化、肩关节囊腋囊增厚对冻结肩的诊断作用，但是日常工作中，肩疼患者常规磁共振检查是主要的，文献缺乏常规磁共振研究肩袖间隙、肩关节囊腋囊（axillary recess，AR）信号强度改变和冻结肩发病的相关性报道。而且，由于肩关节及周围结构众多、彼此关系复杂，影像学上识别较难，其详细的断层解剖及MRI解剖资料较少，尤其缺乏RI薄层断面解剖与磁共振影像学对照研究的资料，而且冻结肩属于I型肩袖间隙病变，因此探讨肩袖间隙断层影像解剖特征对于肩袖间隙疾病诊断、治疗至关重要。研究目的本课题拟基于中国数字化人体(CVH)肩关节数据集和MRI对照，探讨肩袖间隙内重要结构薄层连续断面影像解剖特征，探讨测量喙肱韧带厚度的常规磁共振扫描体位，探讨肩袖间隙脂肪分布、喙肱韧带显示率及厚度与冻结肩发病的相关性，探讨肩袖间隙、肩关节囊腋囊信号强度改变和冻结肩发病的相关性，旨在为肩袖间隙疾病诊断和治疗提供内容详实的薄层断面影像解剖学资料；回答冻结肩是否有MRI征象这个问题，为冻结肩常规磁共振诊断、临床分期、疗效评价提供可参考的客观影像学依据。材料1.中国数字化人体(CVH)肩关节数据集5例，正常健康志愿者20侧肩，磁共振肩关节造影未见异常者5侧肩；1.5T磁共振扫描仪；图形工作站。2.对照组正常健康志愿者60例，共120侧肩，平均年龄50.5岁。冻结肩组单侧冻结肩患者72例72侧肩，平均年龄53.5岁；1.5T磁共振扫描仪。3.对照组正常健康志愿者60例，共120侧肩，平均年龄50.5岁。冻结肩组单侧冻结肩患者120例120侧肩，平均年龄53.3岁；1.5T磁共振扫描仪。方法1.CVH薄层连续断面逐一用肉眼观察，采集后的图像用Photoshop CS2软件处理及标注，并与之相应的MRI肩袖间隙图像进行对照。2.观察指标：(1)肩袖间隙脂肪分布类型：分A，B，C，D，E五型。(2)CHL显示率。(3)CHL厚度。用卡方检验分析肩袖间隙脂肪分布类型、喙肱韧带显示率；用双向方差分析处理喙肱韧带厚度在对照组、冻结肩之间关系及不同性别、不同肩侧别组间的关系。 P＜0.05为差异有统计学意义。3.靶部位信号强度指标：(1)轴位肩袖间隙信号强度SItra；(2)轴位喙突骨皮质SItra；(3)斜矢状位肩袖间隙的SIsag；(4)斜矢状位喙突骨皮质SIsag；(5)斜冠状位肩袖间隙的SIcor；(6)斜冠状位喙突骨皮质SIcor；(7)轴位腋囊的SItra；(8)轴位肱骨皮质SItra；(9)斜冠状位腋囊的SIcor；(10)斜冠状位肱骨皮质SIcor。相对信号强度指标：(1)轴位肩袖间隙的RSItra；(2)斜矢状位肩袖间隙的RSIsag；(3)斜冠状位肩袖间隙的RSIcor；(4)轴位腋囊的RSItra；(5)斜冠状位腋囊的RSIcor。用双向方差分析处理冻结肩、对照组靶部位信号强度、相对信号强度与不同性别、不同肩侧别组间的关系；单向方差分析处理轴位、斜矢状位、斜冠状位肩袖间隙相对信号强度；独立样本t检验处理腋囊相对信号强度。P＜0.05为差异有统计学意义。结果1.肩袖间隙下界SSC及内界CP在轴位、斜矢状位、斜冠状位均清楚、直观；CHL在常规磁共振斜冠状位、斜矢状位最易观察，而轴显示率低；SGHL在CVH显示最清楚，在CVH及MRA上横断面(轴位) SGHL清楚；LBT在CVH上显示最好，可以显示LBT全程（关节内段），其次MRA的斜矢状位；RIC在CVH（尤其横断面）显示好，而MRA及T1WI均不能显示；CVH及MRA斜矢状位显示滑轮结构较好；SGHL、LBT、CP在CVH轴横断面中大致平行，CHL和LBT横断面大致垂直，在LBT前方形成呈“T”分布关系。CVH腋囊显示清楚，为认识该结构提供了可能，但是区分前束、后束困难；MRA显示腋囊厚度困难；压脂PDWI轴位、斜冠状位显示AR呈低信号，优于T1WI。2.(1)肩袖间隙分布类型：冻结肩组和对照组比较，A型显著减少，B型及E型显著增多，卡方检验P＜0.05。(2)CHL显示率：冻结肩斜矢状位CHL显示率(79.2%)低于对照组(91.7%)，而且女性(80%)低于对照组(95%)，卡方检验P＜0.05。冻结肩斜冠状位CHL显示率(87.5%)和对照组(86.7%)比较，卡方检验P＞0.05；冻结肩轴位CHL显示率(19.4%)和对照组(24.2%)比较，卡方检验P＞0.05。(3) CHL厚度：冻结肩斜矢状位CHL厚度3.99±1.68mm(n=57)，大于对照组3.08±1.32mm(n=110)，P＜0.05。冻结肩斜冠状位CHL厚度4.37±1.71mm(n=63)，大于对照组2.84±0.79mm(n=104)，且女性CHL厚度肩大于男性肩，P＜0.05。冻结肩组轴位CHL厚度3.93±1.49mm(n=14)，大于对照组2.29±0.65mm(n=29)，P＜0.05。3.轴位及斜矢状位肩袖间隙信号强度SItra、RSItra，冻结肩大于对照组，P＜0.001，而且女性大于男性，P＜0.05。斜矢状位肩袖间隙相对信号强度RSItra右肩大于左肩，P＜0.05。斜冠状位肩袖间隙信号强度SIcor，RSIcor，冻结肩大于对照组，P＜0.001。轴位、斜冠状位AR信号强度SItra、RSItra，冻结肩大于对照组，P＜0.001，而且女性大于男性，P＜0.05。而且轴位AR信号强度SItra、RSItra,右肩大于左肩, P＜0.05。轴位喙突骨皮质信号强度SItra，冻结肩大于对照组，P＜0.001；斜矢状位喙突骨皮质信号强度SIsag冻结肩大于对照组，而且左肩大于右肩，P＜0.05；斜冠状位喙突皮质信号强度SIcor，冻结肩大于对照组，P＜0.05。轴位、斜冠状位肱骨皮质SI，冻结肩大于对照组，P＜0.001结论1.肩胛下肌腱及喙突可作为磁共振影像学研究肩袖间隙解剖及病变的重要定位标志。CVH和MRI在显示肩袖间隙内容物方面有较好互补性。斜矢状位有利于肩袖间隙边界、内容物及滑轮结构磁共振扫描定位及阅片识别。CVH横断面（轴位）中盂肱上韧带（SGHL）、肱二头肌长头腱（LBT）、喙突大致平行，CHL和LBT大致垂直；CVH及MRA斜矢状位SGHL与CHL大致垂直，在LBT前方形成呈“T”分布关系，明确这些结构的重要空间位置关系，有利于这些解剖结构磁共振扫描定位及阅片识别。MRI压脂PDWI轴位、斜冠状位腋囊显示清楚，为研究肩关节囊解剖及疾病提供了较好的MRI扫描序列及体位。2.肩袖间隙分布类型、CHL显示率、CHL厚度与冻结肩发病有较好的相关性，冻结肩有磁共振征象：冻结肩的肩袖间隙脂肪分布A型减少，B型及E型显著增多；冻结肩CHL显示率下降，其CHL厚度测量以斜冠状位最佳；冻结肩CHL增厚，斜冠状位CHL厚度均值4.37mm及斜矢状位CHL厚度均值3.99mm提示冻结肩。3.肩袖间隙、腋囊压脂PDWI呈高信号改变和冻结肩发病有较好相关性，为冻结肩常规磁共振诊断及阅片提供了靶部位及客观影像学依据：磁共振常规扫描能提示冻结肩诊断，而且阅片重点感兴趣区顺序为：轴位腋囊＞斜冠状位腋囊＞斜矢状位肩袖间隙＞轴位肩袖间隙＞斜冠状位肩袖间隙，即腋囊是优先阅片靶部位。喙突、肱骨近段也可能是冻结肩的受累部位。更多还原

【Abstract】 BackgroundFrozen shoulder, also known as adhesive capsulitis, is an inflammatory condition ofglenohumeral joint capsule and synovium leading to restricted range of motion. The currentconsensus definition of the American Shoulder and Elbow Surgeons is “a condition ofuncertain etiology characterized by significant restriction of both active and passiveshoulder motion that occurs in the absence of a known intrinsic shoulder disorder”. Frozenshoulder most commonly affects women aged between40and60years. In China, frozenshoulder is called the “50s shoulder’’. Frozen shoulder, as well as adhesive capsulitis, iscommonly accepted in the oversea literature.Frozen shoulder is primarily a clinical diagnosis, and it is usually diagnosed on thebasis of clinical findings including history and physical examination. But we poorly graspof the etiology and pathophysiology of frozen shoulder, and initial diagnosis andsubsequent treatment of frozen shoulder is a poorly correct entity today. Invasivearthroscopic findings as the reference “gold standard” for the diagnosis of frozen shoulderis not generally used in routine work. Hence, noninvasive imaging is more and moreimportant in the diagnosis of frozen shoulder.Recently, a number of publications have described the imaging assessment of thediagnosis of frozen shoulder, Some reports have associated osteopenia of the proximalhumerus with frozen shoulder. The glenohumeral joint capacity is less than10ml, whichcan be an indicator of frozen shoulder, using arthrography of the shoulder. Thickenedcoracohumeral ligament can be suggestive of frozen shoulder, using ultrasound. Thepresence of a hypoechoic region with increased vascularity in the rotator interval correlateswith the fibrovascular inflammatory tissue that is usually present and can provide early andaccurate diagnosis of frozen shoulder.Currently, plain magnetic resonance imaging (MRI) and MR arthrography can provide reliable imaging indicators of frozen shoulder, and these findings on MRI have been shownto correlate well with surgical findings, although some studies in the radiologic literature donot support this. Joint capsule and synovium thickness greater than4mm is a useful MRcriterion for the diagnosis of frozen shoulder. Some reports showed that thickness ofcapsule and synovium of the axillary recess greater than3mm is a practical MR criterionfor the diagnosis of frozen shoulder when measured on oblique coronal T2-weighted MRarthrography images without fat suppression. Some reports indicated that blood flow to thesynovium involving the glenohumeral joint increased in the frozen shoulder. The presenceof enhancing fibrovascular scar tissue in the rotator interval(RI), soft tissue thickeningaround the biceps anchor and thickening of the axillary pouch on MR imaging are signssuggestive of frozen shoulder. Some reports showed that thickening of the joint capsule andsynovium and diminished filling ratio of the axillary recess to posterior joint cavityappeared to be useful criterion for the diagnosis of frozen shoulder. Some reports showedthat thickening of the CHL and the joint capsule in the rotator interval, as well as thesubcoracoid triangle sign,were characteristic MR arthrographic findings in frozen shoulder.Some reports showed that there were statistically significant differences in RI dimensionsincluding height, base, RI index, and RI ratio between patients with and without frozenshoulder. Estimating the dimensions of the RI in frozen shoulder using MR arthrographymay prove to be valuable for assessing patients with frozen shoulder preoperatively. Somereports showed that coracohumeral ligament (CHL) thickness on MR arthrographycorrelates with the range limitation of external rotation and internal rotation IR in patientswith frozen shoulder.In the publications, the glenohumeral joint capsule, the rotator interval and thecoracohumeral ligament became the key region of interest of the studies on the frozenshoulder with magnetic resonance imaging (MRI). The thicken CHL was useful for thediagnosis of frozen shoulder, but the measurement of the CHL thickness depended on theCHL visualization rate, and that rate depended on the fat in the rotator interval displayingheperintensity with MRI. Hence, MRI evaluation of the fat distribution type in the rotatorinterval was very important to diagnose the frozen shoulder. However, few studies haveattempted to investigate the fat distribution in the rotator interval, the CHL visualizationrate and the CHL thickness in normal shoulders, using routine magnetic resonance imaging (MRI). There were no reports on the fat distribution in the rotator interval correlated withfrozen shoulder. Some studies showed that the presence of enhancing fibrovascular scartissue in the rotator interval and thickened joint capsule and synovium was useful for thediagnosis of frozen shoulder.However, routine magnetic resonance imaging (MRI) ismainly used in the patients with shoulder pain, there were no reports on the signal intensityof the rotator interval and the axillary recess correlated with frozen shoulder, using routinemagnetic resonance imaging (MRI).Furthermore, frozen shoulder, also known as type Ⅰof RI lesions, was confused withthe other types of the RI lesions because of the poorly understood anatomy of the RI.Though the structures that define the borders of the RI were well agreed upon in theliterature, there remained some controversy as to the exact components comprising thecapsule that bridges this space and their degree of contribution. There were few correlativestudies between thin-sectional anatomy and magnetic resonance imaging (MRI). It wasimportant to explore the morphological features of sectional anatomy of the the RI for thediagnosis and therapy of frozen shoulder.ObjectiveThe current study started from correlative study between thin-sectional anatomy andmagnetic resonance imaging (MRI) based on the subdataset of shoulder from ChineseVisible Human(CVH), to explore the morphological features of continual thin-sectionalanatomy of the important components in the RI. Then, the current study explored therelationship between the fat distribution in the rotator interval, the CHL visualization rateand CHL thickness, using routine magnetic resonance imaging (MRI) to determine the scanplane for measuring the CHL thickness. Finally, The present study investigated thecorrelation between the fat distribution in the rotator interval, the CHL visualization rateand CHL thickness and frozen shoulder, and investigated the correlation between the signalintensity of the rotator interval and the axillary recess and frozen shoulder, using routinemagnetic resonance imaging (MRI). The objective of the present study was provide thedetailed thin-sectional anatomy of the RI for the diagnosis and treatment of the RI lesions,and to investigate whether the patients with frozen shoulder had some useful signs withroutine MRI. The objective of the present study was also to supply the magnetic resonanceimaging features for diagnosing frozen shoulder, evaluating the stages and therapy of frozen shoulder.Materials1.The subdataset of the shoulder from Chinese Visible Human(CVH)Number1toNumber5,20shoulder joints of the normal volunteer individuals,5shoulders withoutabnormalities from5patients underwent MR arthrography, a1.5-T system, and an imageworkstation with the visualizational software were included to use in the first part of thepresent study.2. One hundred and twenty shoulder joints in60normal volunteer individuals (a meanage of50.5years),72shoulder joints in72patients(a mean age of53.5years)with frozenshoulder, and a1.5-T system were included to use in the second part of the present srudy.3. One hundred and twenty shoulder joints in60normal volunteer individuals (a meanage of50.5years),120shoulder joints in120patients (a mean age of53.3years) withfrozen shoulder, and a1.5-T system were included to use in the third part of the presentsrudy.Methods1.The subdataset of shoulder from CVH was observed to compare the detailedsectional anatomy structure of the RI with routine MRI and MR arthrography. The anatomystructure of the RI and its components was marked from CVH, routine MRI and MRarthrography one by one with Photoshop CS2software.2.A MRI evaluation index included the fat distribution type, the rate of CHLvisualization, and the CHL thickness. The fat distribution types included type A, type B,type C, type D and type E. A chi-square test was used to analyze the data for the fatdistribution type and the rate of CHL visualization. A two-way ANOVA was used to analyzethe maximum thickness of CHL for different lateral shoulders and different gendershoulders.Two-tailed hypothesis tests were used, and local statistical significance wasassumed to be P <0.05for all parameters.3.Evaluation index included the signal intensity (SI)of the RI(the SItra, the SIsag,andthe SIcor of the RI), the SI of the cortical bone of coracoid process(CP)(the SItra, theSIsag,and the SIcor of the cortical bone of CP), the SI of the axillary recess(AR)(theSItra and the SIcor of the AR), and the SI of the cortical bone of humerus (the SItra and theSIcor of the cortical bone of humerus). Evaluation Indicator also included the relative SItra (RSItra) of the RI, the RSIsag of the RI, the RSIcor of the RI, the RSItra of the axillaryrecess(AR), and the RSIcor of the AR. A two-way ANOVA was used to analyze theevaluation index for different lateral shoulders and different gender shoulders. A one-wayANOVA was used to analyze the the RSItra of the RI, the RSIsag of the RI and the RSIcorof the RI. An independent-samples t test was used to analyze the RSItra of the AR and theRSIcor of the AR. Two-tailed hypothesis tests were used, and local statistical significancewas assumed to be P <0.05for all parameters.Results1.The inferior border(the superior margin of SSC)of the RI and the medial border(CP))of RI were markedly displayed on transverse, sagittal oblique and coronal obliqueplane. The CHL was markedly displayed on the sagittal oblique and coronal oblique plane,using plain MRI, and the rate of the CHL was low on the transverse plane.The SGHL wasmarkedly displayed on the CVH, especially in the transverse plane. The LBT was markedlydisplayed on the CVH, and the intraarticular full segment of the LBT was also markedlydisplayed on the CVH. The LBT was markedly displayed on the sagittal oblique plane ofMR arthrography. The RIC was markedly displayed on the CVH, and was not displayedwith plain MRI and MR arthrography. The biceps reflection pulley was markedly displayedon the sagittal oblique plane with CVH and MR arthrography. The SGHL and the LBT wereparallel to the coracoid process on the transverse plane from CVH. The CHL wasperpendicular to the LBT in the transverse plane from CVH. The CHL was perpendicular tothe SGHL, with T-shaped link anterior to the LBT on the sagittal oblique plane.The axillaryrecess (AR) was markedly displayed on transverse plane from the subdataset of theshoulder of CVH, but the AR was difficult to differentiate between the anterior band andthe posterior band of the inferior GHL complex. The thickness of the glenohumeral jointcapsule in the AR was too thinner to be measured in fat-suppressed T1WI from MRarthrography.The glenohumeral joint capsule in the AR displayed low signal intensity in thefat-suppressed PDWI, and was markedly displayed in the transverse and coronal obliqueplanes.2.(1)A comparison of the fat distribution types in the RI in the patients with frozenshoulder versus the control group showed that the number of the type A decreased, thenumber of the type B and type E increased, and there was significant difference, using the x2test (P＜0.05).(2)A comparison of the CHL visualization rate in the patients with frozenshoulder versus the control group showed that the CHL visualization rate (91.7%) in thecontrol group was significantly greater than that rate in the frozen shoulder group (79.2%),and the CHL visualization rate(95%) in the female shoulder from the control group wassignificantly greater than that rate(80%) in the female shoulder with frozen shoulder (P＜0.05). The CHL visualization rate in the coronal oblique images was86.7%in the controlgroup, and87.5%in the frozen shoulder group, and there was no significant difference (P＞0.05). The CHL visualization rate in the transverse images was24.2%in the controlgroup, and19.4%in the frozen shoulder group, there was no significant difference (P＞0.05).(3)A comparison of the CHL thickness in the patients with frozen shoulder versus thecontrol group showed that the CHL thickness (3.99±1.68mm)on sagittal oblique plane inthe patients with frozen shoulder was significantly greater than that thickness (3.08±1.32mm) for the control group, using a two-way ANOVA (P＜0.05). The CHL thickness(4.37±1.71mm)on coronal oblique plane in the patients with frozen shoulder wassignificantly greater than that thickness (2.84±0.79mm) for the control group(P＜0.05),and the CHL thickness on the coronal oblique plane in the female shoulders wassignificantly greater than that thickness in the male shoulders (P＜0.05). The CHLthickness (3.93±1.49mm)on transverse plane in the patients with frozen shoulder wassignificantly greater than that thickness (2.29±0.65mm) for the control group (P＜0.05).3.A comparison of the SItra and the RSItra of the RI in the patients with frozenshoulder versus the control group was significant, using a two-way ANOVA(P＜0.001),andthe SItra and the RSItra of the RI in the female shoulders was significantly greater than thatSItra and RSItra in the male shoulders (P＜0.05). The SIsag and the RSIsag of the RI wasthe same result as that SItra and the RSItra of the RI between the patients and the controlgroup (P＜0.001), and the RSIsag of the RI in the right shoulders was significantly greaterthan that RSIsag of the RI in the left shoulders (P＜0.05). A comparison of the SIcor andthe RSIcor of the RI in the patients with frozen shoulder versus the control group wassignificant(P＜0.001). A comparison of the SItra，the RSItra the AR in the patients withfrozen shoulder versus the control group was significant (P＜0.001), and the SIcor andthe RSIcor of the AR was the the same result as that SItra and the RSItra of the AR betweenthe patients and the control group (P＜0.001). The SItra and the RSItra of the AR in the female shoulders was significantly greater than that SItra and that RSItra of the AR in themale shoulders (P＜0.05). The SItra and the RSItra of the AR in the right shoulders wassignificantly greater than that SItra and of that RSItra the AR in the left shoulders (P＜0.05).A comparison of the SItra, the SIsag and the SIcorof the cortical bone of CP in the patientswith frozen shoulder versus the control groupwas significant(P＜0.001),and the SIsag ofthe cortical bone of CP in the left shoulderswas significantly greater than that SIsag in theright shoulders (P＜0.05). A comparison of the SItra and the SIcor of the cortical bone ofhumerus in the patients with frozen shoulder versus the control group was significant.Conclusions1.The SSC and the CP is important positioning mark for evaluating of the anatomy andlesions of the RI,using MRI. It is complementary for MRI and CVH displaying thecomponents of the RI. The sagittal oblique is the best position for displaying the borders ofthe RI, the components of the RI and the biceps reflection pulley, and that plane was usefulfor the scanning-position mark and reading the film of MRI. The SGHL and the LBT wereparallel to the coracoid process in the transverse plane from CVH.The CHL wasperpendicular to the LBT in the transverse plane from CVH. The CHL was perpendicular tothe SGHL, with T-shaped link anterior to the LBT on the sagittal oblique plane from CVHand MR arthrography. That spatial relationship among those components of the RI wasuseful for the scanning-position mark and reading the film of MRI. The glenohumeral jointcapsule in the AR displayed low signal intensity in the fat-suppressed PDWI, and wasmarkedly displayed in the transverse and coronal oblique planes for the anatomy andlesions of the glenohumeral joint capsule.2. The change of the fat distribution type in the RI, the rate of CHL visualization andthe thickened CHL correlates with frozen shoulder. There is a few of imaging signs of thefrozen shoulder with routine MRI.The number of the type A for the fat distribution in theRI decreases, and the number of the type B and type E increases in the patients with frozenshoulder. The rate of CHL visualization decreases in the patients with frozen shoulder.Thecoronal oblique plane is the best to the CHL depiction and measuring the CHL thickness inthe patients with frozen shoulder. A thickened CHL in the coronal oblique plane(4.37mm)and in the sagittal oblique plane (3.99mm) is highly suggestive of frozen shoulder, perhapstaken as one of the most characteristic MR findings for frozen shoulder. 3.The hyperintensity of the rotator interval and the axillary recess in fat-suppressedproton-density weighted spin-echo images(PDWI) correlates with frozen shoulder, usingroutine magnetic resonance imaging (MRI). Routine MRI is useful for the diagnosis offrozen shoulder, and the key regions of interest of lesions with frozen shoulder in thefat-suppressed PDWI should be orbserved from the order: at first, the AR in the transverseplane should be orbserved. then, the AR in the coronal oblique plane,the RI in the sagittaloblique plane and the RI in the transverse plane. at last, the RI in the coronal oblique planeshould be orbserved. Of course, the affected area of frozen shoulder may include theproximal humerus and the CP.更多还原