New MRI Grading System for the Cervical Canal Stenosis

Transcription

1 Musculoskeletal Imaging Original Research Kang et al. MRI Grading of Cervical Canal Stenosis Musculoskeletal Imaging Original Research Yusuhn Kang 1 Joon Woo Lee 1 Young Hwan Koh 2 Saebeom Hur 3 Su Jin Kim 1 Jee Won Chai 2 Heung Sik Kang 1 Kang Y, Lee JW, Koh YH, et al. Keywords: central canal stenosis, cervical spine, grading, sagittal MRI DOI: /AJR Received August 19, 2010; accepted after revision December 1, Department of Radiology, Seoul National University Bundang Hospital, 300 Gumidong, Bundag-Gu, Seong Nam, Gyeongi-do , Republic of Korea. Address correspondence to J. W. Lee (joonwoo2@gmail.com). 2 SMG-SNU Borame Medical Center, Seoul, Republic of Korea. 3 Department of Radiology, Seoul National University Hospital, Seoul, Republic of Korea. WEB This is a Web exclusive article. AJR 2011; 197:W134 W X/11/1971 W134 American Roentgen Ray Society New MRI Grading System for the Cervical Canal Stenosis OBJECTIVE. The purpose of this study was to propose a new MRI grading system for cervical canal stenosis and to evaluate the reproducibility of the system. MATERIALS AND METHODS. Cervical canal stenosis was classified according to the T2-weighted sagittal images into the following grades: grade 0, absence of canal stenosis; grade 1, subarachnoid space obliteration exceeding 50%; grade 2, spinal cord deformity; and grade 3, spinal cord signal change. The MRI scans of 82 patients (37 men and 45 women; mean age, 65.2 years; range, years) were independently analyzed by six radiologists. Interobserver and intraobserver agreements were analyzed using intraclass correlation coefficient (ICC), along with the percentage agreement and kappa statistics. RESULTS. The ICC for interobserver agreement was , indicating good-toexcellent agreement. For the distinction among the four grades, the percentage of agreement was 63 64% (κ = ). The percentage of agreement for the presence of cervical canal stenosis (grade 0 vs grades 1, 2, and 3) was 79 85% (κ = ). The percentage of agreement for insignificant (grade 0 1) or significant (grade 2 3) stenosis was 81 85% (κ = ). The percentage of agreement for the presence of spinal cord signal change (grade 0 2 vs grade 3) was 92 95% (κ = ). The overall intraobserver agreement was excellent, as determined by an ICC of CONCLUSION. The new grading system provides a reliable assessment of cervical canal stenosis. C ervical spondyliotic myelopathy is a common degenerative disorder associated with the narrowing, or stenosis, of the spinal canal that frequently affects elderly patients. Narrowing of the spinal canal may be caused by various factors, including the herniation or bulging of intervertebral disks, osteophytes, and ossification of the posterior longitudinal ligament. The progressive compression of the cord by these factors may lead to spinal cord ischemia, resulting in histopathologic changes of the cervical spinal cord [1, 2]. Previous studies have described various methods of assessing the degree of cervical canal stenosis. Early studies were based on radiographs; Edwards and Larocca [3] measured the sagittal diameter of the cervical spinal canal on a plain lateral radiograph, whereas Pavlov et al. [4] and Torg et al. [5] used the ratio of the sagittal diameter of the cervical canal divided by the corresponding diameter of the vertebral body. However, MRI is currently by far the most commonly used imaging method for the accurate evaluation of spinal canal stenosis. MRI visualizes not only the width and length of the spinal canal but also depicts in detail the spinal cord, intervertebral disks, osteophytes, and ligaments, all of which are potential causes of spinal canal stenosis [2]. Despite the various assessments made in previous studies, grading systems for cervical canal stenosis based on MRI and corresponding reproducibility studies are currently sparse. Moreover, there is no universally used grading system. A standardized grading system in the assessment of cervical canal stenosis is a prerequisite for the comparison of data from different investigations and for the improvement of communication between radiologists and clinicians. Muhle et al. [6] classified cervical canal stenosis according to the following grading system: grade 0, normal; grade 1, partial obliteration of the anterior or posterior subarachnoid W134 AJR:197, July 2011

2 MRI Grading of Cervical Canal Stenosis space; grade 2, complete obliteration of the anterior or posterior subarachnoid space; and grade 3, cervical cord compression or displacement. In our hospital, we have adapted the grading system proposed by Muhle et al. in reporting cervical canal stenosis. In our experience over the last few years, there are several limitations to the grading system of Muhle et al. First, the definition of partial obliteration is unclear. Second, we encountered cases in which CSF clefts were notable around the spinal cord, even when there was severe canal compromise or spinal cord deformity. In turn, the definition of complete obliteration is also impractical. Third, no consideration is given to the signal change of the spinal cord, which is known as a good sign of compressive myelopathy [7 10]. Therefore, we made a new grading system for diagnosing and grading cervical canal stenosis that is based on the preexisting grading system by Muhle et al. [6]. The purpose of this study was to propose an MRI grading system for cervical canal stenosis and to evaluate its reproducibility. Materials and Methods MRI Grading System for Cervical Canal Stenosis An MRI grading system for cervical canal stenosis was developed by two experienced musculoskeletal radiologists in consensus, on the basis of the literature and previously published work [6, 11 13]. Cervical canal stenosis was classified into a grading system according to the MRI findings on T2-weighted sagittal images (Fig. 1). Grade 0 refers to the absence of central canal stenosis (Fig. 2). Grade 1 refers to nearly complete obliteration of subarachnoid space, including obliteration of the arbitrary subarachnoid space exceeding 50%, without signs of cord deformity (Fig. 3). Grade 2 refers to central canal stenosis with cord deformity but without spinal cord signal change (Fig. 4). Grade 3 refers to the presence of spinal cord signal change near the compressed level on T2-weighted images (Fig. 5). Fig. 1 Schematic diagrams of grading system of cervical canal stenosis in sagittal scans of cervical spines. Grade 0 is normal. Grade 1 denotes obliteration of more than 50% of subarachnoid space without any sign of cord deformity. Grade 2 denotes central canal stenosis with spinal cord deformity; cord is deformed but no signal change is noted in spinal cord. Grade 3 denotes increased signal intensity of spinal cord near compressed level on T2- weighted images. Fig year-old man without cervical canal stenosis. Sagittal T2-weighted fast spin-echo image shows cervical spine without compromise of spinal canal. Normal CSF space is visible around spinal cord, and there is no evidence of cord deformity or signal change within cord. Case Selection This study was approved by our institutional review board, and informed consent was not required. This retrospective study involved cervical spinal MRI scans of 100 patients older than 60 years. The age range of 60 years or older was selected on the basis of an anatomic study in cadavers by Lee et al. [14], which showed that specimens from donors who were 60 years old or older at the time of death had significantly narrower canals than specimens from younger donors. This helped to eliminate a majority of MRI scans with negative findings. Fifty patients who had undergone cervical spinal MRI in our institution between the period of January and September 2009 were consecutively selected. To include MRI scans from various scanners, another 50 patients with cervical spinal MRI scans imported from outside hospitals were also consecutively selected. Among the selected 100 patients, 18 patients with either history of trauma, previous spinal operation, or histologically proven or suspected tumor were excluded on the basis of medical records, as reviewed by one of the authors. As a result, 82 patients were included in the final analysis (37 men and 45 women; mean age, 65.2 years; range, years). Fig year-old man with cervical canal stenosis. Sagittal T2-weighted fast spin-echo image shows grade 1 stenosis with obliteration of CSF space exceeding 50% of arbitrary subarachnoid space at C4 5, C5 6, and C6 7 levels (arrows). AJR:197, July 2011 W135

3 Kang et al. consensus grade was allocated to each level. To assess the intraobserver variability, one reader interpreted the images twice with an interval of 3 months, to minimize observer memory bias. Fig year-old woman with cervical canal stenosis. Sagittal T2-weighted fast spin-echo image shows grade 2 stenosis at C4 5 level (arrow). Spinal cord is compressed and deformed, but spinal cord shows no signal changes. Grade 1 stenoses were also seen at C5 6 and C6 7 levels. Image Acquisition In our institute, the MRI scans were performed with either a 1.5-T scanner (Gyroscan, Philips Healthcare) or 3-T scanner (Intera Achieva, Philips Healthcare) using a 16-channel neurovascular coil for the 3-T scanner or a head-and-neck coil for the 1.5-T scanner, with the patient in the supine position. The imaging protocol included sagittal T2-weighted images acquired with the following parameters: TR/TE, / ; field of view, mm; and section thickness, mm. The lower end of the range of TR values would yield significant T1 weighting. However, the contrast and sharpness of the subarachnoid space spinal cord interface was sufficient for grading, and the T1 weighting on these images was presumed to have insignificant influence on the grading of stenosis. The imported MRI scans performed at diverse hospitals before the referral to our hospital were performed with various MRI scanners using different protocols. T2-weighted sagittal images were selected for grading. All the images were deidentified Fig year-old woman with cervical canal stenosis. Sagittal T2-weighted fast spin-echo image shows grade 3 cervical canal stenosis at C5 6 level (arrow). Spinal canal is significantly narrow at C5 6 level, and signal intensity of spinal cord is increased at corresponding level. before the analysis and were randomly numbered from 1 to 82. Image Interpretation Six observers with different levels of experience two radiology residents, one radiology fellow with 1 year of training in the musculoskeletal subspecialty, and three experienced musculoskeletal radiologists with 10, 7, and 3 years of experience, respectively independently interpreted the images, blinded to the patient s history and age. The images were given in the form of a presentation slide, as a set of three images for each case, and included the median sagittal plane image and the two images bilaterally adjacent to the median plane. Display monitors were not standardized across readers. The grading system was introduced to the readers as a written description, as specified earlier in this article, without pictorial or diagrammatic examples. The observers were asked to rate the degree of cervical canal stenosis according to the aforementioned grading system at three sequential levels (C4 5 to C6 7), and a Evaluation of Patient Symptoms The medical records of the patients were reviewed by one author, focusing on the following points: first, the presence of severe pain, which was defined as 5 points or more measured on an 11-point pain intensity numeric rating scale, where 0 denotes no pain and 10 denotes the worst possible pain; second, the presence of neurologic deterioration suggestive of myelopathy, including motor weakness of the upper or lower extremity and sensory symptoms such as paresthesia [2]; and third, whether the patient had undergone surgery for cervical canal stenosis or had been considered as a surgical candidate by surgeons after the MRI was performed. The patients were grouped according to the highest grade of stenosis among the three levels, with reference to the consensus grade. Each classified group was analyzed for the presence of the aforementioned points. Statistical Analysis Intraobserver and interobserver reliability of the new MRI grading system were assessed by using intraclass correlation coefficient (ICC), with the two-way random or one-way random effects model. Among the various guidelines for the interpretation of ICC, we have adapted a scale introduced in previous literature: ICC values of less than 0.40 indicate poor reproducibility, ICC values of indicate fair-to-good reproducibility, and ICC values greater than 0.75 indicate excellent reproducibility [15, 16]. The percentage of agreement and kappa statistics were also calculated for the distinction among the four individual grades, the distinction between insignificant and significant stenosis (grades 0 and 1 vs grades 2 and 3), the presence of central canal stenosis (grade 0 vs grades 1, 2, and 3), and the presence of signal change in the spinal cord (grades 0, 1, and 2 vs grade 3). As for the distinction among the four individual grades, agreement was measured using the linear weighted Cohen kappa statistic, giving greater weight to a difference of more than one step between observers than to a difference of only one step [17]. The percentage agreement and kappa statistic were assessed for all 15 reader pairs, and the result was averaged. The positive and negative agreements were calculated along with the overall percentage agreement to address any imbalance in the proportion of positive and negative responses [18]. According to Landis and Koch [19], the agreement was rated as follows: kappa values of W136 AJR:197, July 2011

4 MRI Grading of Cervical Canal Stenosis TABLE 1: Distribution of Stenoses by Cervical Level Cervical Level Grade 0 Grade 1 Grade 2 Grade 3 Total C (42.7) 22 (26.8) 16 (19.5) 9 (11.0) 82 C (25.6) 33 (40.2) 14 (17.1) 14 (17.1) 82 C (30.5) 36 (43.9) 16 (19.5) 5 (6.1) 82 Overall 81 (32.9) 91 (37.0) 46 (18.7) 28 (11.4) 246 Note Data are no. (%) of levels. indicated slight agreement, indicated fair agreement, indicated moderate agreement, indicated substantial agreement, and 0.81 or greater indicated excellent agreement. SPSS (version 16.0, SPSS) and MedCalc software (version , MedCalc Software) were used for statistical analyses. Results The distribution of the stenoses based on the consensus grade is shown on Table 1. Grade 1 stenosis was most frequently noted, and the overall burden of stenosis was greatest at the C5 6 level. Interobserver Agreement Table 2 shows the interobserver agreement with respect to ICC, percentage agreement, and kappa statistics. The ICC ranged from to 0.802, indicating good-to-excellent agreement. The average percentage agreement was 63 95% (κ = ) at the C4 5 level, 64 92% (κ = ) at the C5 6 level, and 63 95% (κ = ) at the C6 7 level, according to how the categories were grouped. For the distinction among the four individual grades, the percentage agreement ranged from 63% to 64%, and the kappa values (κ = ) indicated moderate-to-substantial agreement. A slightly higher level of agreement was seen for the presence of central canal stenosis (grade 0 vs grades 1, 2, and 3; percentage agreement, 79 85%; κ = ). Agreement was even higher for the classification of cases as either insignificant stenosis (grades 0 and 1) or significant stenosis (grades 2 and 3); the percentage agreement was 81 85%, and kappa values were , indicating substantial agreement. The best agreement was seen for the presence of signal change in the spinal cord (grades 0, 1, and 2 vs grade 3; percentage agreement, 92 95%; κ = ). The average agreement among the three more experienced radiologists was also calculated, and the results did not greatly differ from the average agreement of all six readers (also shown in Table 2). Intraobserver Agreement The overall intraobserver agreement was excellent, as determined by an ICC value of (Table 3). The results of the analysis of intraobserver reliability based on percentage agreement and kappa statistics are also summarized in Table 3. The percentage agreement was 67 93% (κ = ) at the C4 5 level, 68 91% (κ = ) at the C5 6 level, and 76 98% (κ = ) at the C6 7 level, on the basis of how the categories were grouped. The level of agreement was similar to that of the interobserver study. Evaluation of Patient Symptoms Because of the retrospective nature of this study, there were missing data elements in the medical records, which may have limited our TABLE 2: Interobserver Reliability All Readers (n = 6) Experienced Readers (n = 3) Grouping, Cervical Level ICC Agreement (%) Weighted κ ICC Agreement (%) Weighted κ Grading C C C Overall Stenosis (grade 0 vs grades 1 3) C C C Overall Significance (grades 0 1 vs grades 2 3) C C C Overall Signal change (grades 0 2 vs grade 3) C C C Overall Note Dashes indicate not accessible. ICC = intraclass correlation coefficient. AJR:197, July 2011 W137

5 Kang et al. evaluation. Among the 52 patients who complained of pain, its severity was not specified in the medical records for 25 patients. Although the presence of severe pain did not show correlation with the grade of cervical canal stenosis, the percentage of patients with neurologic deterioration suggestive of myelopathy increased with the grade of stenosis. The proportion of patients who had either undergone surgery or were considered as surgical candidates also increased with increasing stenosis grade. None of the patients with grade 0 stenosis was considered as a surgical candidate. The results are summarized in Table 4. TABLE 3: Intraobserver Reliability Grouping, Cervical Level ICC Agreement (%) Weighted κ Grading C C C Overall Stenosis (grade 0 vs grades 1 3) C C C Overall Significance (grades 0 1 vs grades 2 3) C C C Overall Signal change (grades 0 2 vs grade 3) C C C Overall Note Dashes indicate not accessible. ICC = intraclass correlation coefficient Discussion According to our results, the MRI grading system showed intraobserver and interobserver agreement sufficient enough to serve as a reliable method for evaluating and reporting the degree of cervical canal stenosis. In a clinical setting, determining the presence of spinal canal stenosis or the significance of the stenosis, and also determining whether there are signs of compressive myelopathy, may be more important than distinguishing among the four different grades of stenosis. Our grading system could be further grouped to differentiate between the presence and absence of stenosis, insignificant and significant stenosis, and the presence and absence of cord signal change. In all of these groupings, the intraobserver and interobserver variability was considerably high, suggesting that the grading system is useful in a clinical context. The interobserver variability was assessed among readers with different levels of experience working in different institutions. Considering that only a written description of the grading system was provided to the readers, the percentage of 63 95% is a fairly high level of agreement. This finding implies that the grading system is simple enough to be applied consistently and to be understood and learned. The average agreement of all six readers and that of the three more experienced readers showed little difference, also suggesting that the system is easy to learn, even for the inexperienced radiologist. The interobserver agreement may be further enhanced by providing a conceptual diagram and representative cases of each grade. The intraobserver agreement of this study was assessed only with the readings of one inexperienced observer. At the beginning of the study, we thought the intraobserver agreement of an inexperienced observer would be more important, because a broadly applicable grading system should be easy to learn, even for an inexperienced observer. In addition, on the basis of the results that showed little difference between the average agreement of all six readers and that of the three more experienced readers, we assume that the intraobserver agreement would not be significantly affected by the level of experience of the reader. The grade of stenosis showed a positive correlation with the percentage of patients with neurologic deterioration suggestive of myelopathy and with the proportion of the patients who had either undergone surgery or who were considered as surgical candidates. This finding indicates that our grading system conveys clinically significant information and that it may correlate well with the recommended indications for surgical treatment, although further studies are mandatory. The assessment of cervical canal stenosis based on MRI scans has been attempted in various studies. In a study by Stafira et al. [20], the interobserver agreement among radiologists with regard to the level, degree (none, mild, moderate, or severe), and cause (bone, disk, or combination) of cervical spinal stenosis was evaluated qualitatively, without any guideline or instructions. The observed kappa values for degree and cause of stenosis based on MRI scans were 0.31 and 0.22, respectively. The marked inconsistency among observers indicates the necessity of a uniform standard in the evaluation of stenosis that can be universally accepted and applied. Nagata et al. [11] reported the usefulness of sagittal T1- weighted MRI in diagnosing cervical myelopathy and its good correlation with the severity of cord compression. Compressed deformities were evaluated as class 0 (no compression), class 1 (cord compressed slightly), class 2 (cord width decreased by less than one third), and class 3 (cord width decreased by at least one third). However, this classification by Nagata et al. mainly focused on spinal cord compression, without a definition or grading for cervical canal stenosis. The aforementioned classification of Muhle et al. [6] was based on the obliteration of the subarachnoid space surrounding the spinal cord and the presence of cord compression or displacement. The impractical definition of partial or complete obliteration of the subarachnoid space and the absence of consideration of cord signal change were the limitations of this grading system, as mentioned earlier. A W138 AJR:197, July 2011

6 MRI Grading of Cervical Canal Stenosis TABLE 4: Correlation Between Grade of Stenosis and Patient Symptoms Grade No. of Patients Total b Severe c Unspecified d Deterioration Surgery e a Pain Neurologic (83.3) 3 (50.0) 1 (16.7) 1 (16.7) 0 (0.0) (50.0) 9 (25.0) 7 (19.4) 16 (44.4) 3 (8.3) (70.0) 5 (25.0) 6 (30.0) 13 (65.0) 5 (25.0) (75.0) 4 (20.0) 11 (55.0) 18 (90.0) 19 (95.0) Total Note Data are no. (%) of patients. a Patients are grouped according to the highest grade of stenosis. b Total no. of patients who complained of pain in the neck or upper extremity, without regard to severity. c Pain score of 5 or higher on an 11-point numeric rating scale. d Pain of unspecified severity in the neck or upper extremity. e Patients who have undergone surgery or have been considered as a surgical candidate. simple one-dimensional quantitative assessment of stenosis was proposed by Larsson et al. [12]; mild narrowing referred to less than a 50% reduction in the width of the subarachnoid space, moderate narrowing referred to a greater than 50% reduction in the subarachnoid space, and severe stenosis was defined as cord compression. However, this classification system also lacked the consideration of cord signal change, which is a finding well correlated with cervical spinal myelopathy [7 9]. Despite these efforts, to our knowledge, no universally adopted scheme of classification is in use in the field of radiology or the surgical counterparts. This may hinder the communication between radiologists and clinicians and further hamper the clinical research on cervical canal stenosis and cervical spondyliotic myelopathy. It has been shown in previous studies that the severity of stenosis, which may involve the static factors of cervical myelopathy, is a significant predictive factor for clinical outcome in the elderly patients [21]. However the definition of severe stenosis varies among radiologists and remains ambiguous. There is no doubt that a standardized assessment of cervical canal stenosis is necessary. For an established classification system, the information conveyed by the system must be not only clear cut and simple but also clinically important. Takahashi et al. [7] have shown that the severity of spinal canal distortion with cord compression is directly proportional to the severity of clinical impairment and that the presence of high signal intensity is an indicator of poorer prognosis. A similar finding was postulated in a study by Harrop et al. [13] in which a strong correlation was shown between the presence of cervical myelopathic findings on physical examination and radiographic cervical spinal cord compression and hyperintense T2 intraparenchymal spinal cord signal abnormalities. Our grading system incorporates these two clinically important MRI findings, and they are defined as key findings of grade 2 and 3 stenosis, respectively. An earlier stage of stenosis with obliteration of more than 50% of the arbitrary subarachnoid space, but without any sign of cord deformity or cord signal change, was defined as grade 1. Most patients symptoms deteriorate over the years [22], and there is evidence that about 5% of all patients with asymptomatic spinal cord compression become symptomatic each year [23]. In other words, patients with grade 1 stenosis are prone to progress to higher grades of cervical spinal stenosis and myelopathy. In these terms, it is important to identify these patients and to provide appropriate medical attention. Our study may be limited by several factors. First, the evaluation of the correlation between the stenosis grade and patient symptoms was limited, because of the retrospective nature of our study. A crude correlation between the grade of stenosis and neurologic deterioration suggestive of myelopathy has been shown. However, further investigation based on elaborate assessment of symptoms, such as the Japanese Orthopedic Association score [24], is needed to elucidate the correlation between the grades and the severity of symptoms and the association of preoperative grades with outcome in patients undergoing surgical treatment. Second, the grading system was based on the sagittal T2-weighted images only. We focused on the simplicity of the grading system for its universal use. A study by Ryan et al. [25] has suggested that T2-weighted imaging alone is sufficient to diagnose the cause of cervical myelopathy and radiculopathy and that T1-weighted imaging does not contrib- ute clinically significant information in addition to that provided by the T2-weighted images alone. In our experience, the sagittal T2-weighted image is the single most important sequence for the evaluation of cervical canal stenosis and therefore was selected as the target of grading. Although additional imaging sequences may not be mandatory for the grading of spinal stenosis, we expect that other imaging sequences, including T1- weighted images, will play a complementary role in depicting the cause of spinal stenosis, revealing soft-disk herniation and marrow infiltration. Third, this study was conducted at a single center, which perhaps limits its generalizability. We included MRI scans imported from an outside hospital to partly overcome this limitation. However, further multicenter studies are required. Finally, our study was based on recumbent MRI scans. Studies have shown that the conventional recumbent MRI of the cervical spine may underestimate disease because the imaging is performed in a nondynamic non-weightbearing position [26, 27]. The upright MRI system is currently not available in our institution and, inevitably, recumbent MRI scans were analyzed. However, we think that the application of the new grading system does not necessarily have to be limited to recumbent MRI scans. The grading system may be applicable to upright MRI scans with sufficient interobserver and intraobserver reliability, although further study is warranted. In conclusion, the new grading system provides a reliable assessment of cervical canal stenosis. References 1. Cook C, Braga-Baiak A, Pietrobon R, et al. Observer agreement of spine stenosis on magnetic resonance imaging analysis of patients with cervical AJR:197, July 2011 W139

7 Kang et al. spine myelopathy. J Manipulative Physiol Ther come in cervical spondylitic myelopathy? Spine server variability on CT and MR images. AJNR 2008; 31: (Phila Pa 1976) 1999; 24: ; 24: Bernhardt M, Hynes RA, Blume HW, White AA. 11. Nagata K, Kiyonaga K, Ohashi T, Sagara M, Mi- 21. Morio Y, Teshima R, Nagashima H, Nawata K, Cervical spondylitic myelopathy. J Bone Joint yazaki S, Inoue A. Clinical value of magnetic Yamasaki D, Nanjo Y. Correlation between oper- Surg Am 1993; 75: resonance imaging for cervical myelopathy. Spine ative outcomes of cervical compression myelopa- 3. Edwards WC, Larocca H. The developmental seg- (Phila Pa 1976) 1990; 15: thy and MRI of the spinal cord. Spine (Phila Pa mental sagittal diameter of the cervical spinal canal in patients with cervical spondylosis. Spine (Phila Pa 1976) 1983; 8: Pavlov H, Torg JS, Robie B, Jahre C. Cervical spinal stenosis: determination with vertebral body ratio method. Radiology 1987; 164: Torg J, Pavlov H, Genuario S, et al. Neuropraxis of the cervical spinal cord with transient quadriplegia. J Bone Joint Surg Am 1986; 68: Muhle C, Metzner J, Weinert D, et al. Classification system based on kinematic MR imaging in cervical spondylitic myelopathy. AJNR 1998; 19: Takahashi M, Yamashita Y, Sakamoto Y, Kojima R. Chronic cervical cord compression: clinical significance of increased signal intensity on MR images. Radiology 1989; 173: Yukawa Y, Kato F, Yoshihara H, Yanase M, Ito KMR. T2 image classification in cervical compression myelopathy. Spine (Phila Pa 1976) 2007; 32: Al-Mefty O, Harkey LH, Middleton TH, Smith RR, Fox JL. Myelopathic cervical spondylitic lesions demonstrated by magnetic resonance imaging. J Neurosurg 1988; 68: Wada E, Yonenobu K, Suzuki S, Kanazawa A, Ochi T. Can intramedullary signal change on magnetic resonance imaging predict surgical out- 12. Larsson E-M, Holtås S, Cronqvist S, Brandt L. Comparison of myelography, CT myelography and magnetic resonance imaging in cervical spondylitis and disk herniation: pre- and postoperative findings. Acta Radiol 1989; 30: Harrop JS, Naroji S, Maltenfort M, et al. Cervical myelopathy: a clinical and radiographic evaluation and correlation to cervical spondylitic myelopathy. Spine (Phila Pa 1976) 2010; 35: Lee MJ, Cassinelli EH, Riew KD. Prevalence of cervical spine stenosis: anatomic study in cadavers. J Bone Joint Surg Am 2007; 89: Rosner B. Fundamentals of biostatistics, 6th ed. Belmont, CA: Duxbury Press, Sampat MP, Whitman GJ, Stephens TW, et al. The reliability of measuring physical characteristics of spiculated masses on mammography. Br J Radiol 2006; 79:S134 S Sim J, Wright CC. The kappa statistic in reliability studies: use, interpretation, and sample size requirements. Phys Ther 2005; 85: Kundel HL, Polansky M. Measurement of observer agreement. Radiology 2003; 228: Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: Stafira JS, Sonnad JR, Yuh WTC, et al. Qualitative assessment of cervical spinal stenosis: ob- 1976) 2001; 26: Meyer F, Börm W, Thomé C. Degenerative cervical spinal stenosis: current strategies in diagnosis and treatment. Dtsch Arztebl Int 2008; 105: Bednarik J, Kadanka Z, Dusek L, et al. Presymptomatic spondylitic cervical cord compression. Spine (Phila Pa 1976) 2004; 29: Yonenobu K, Abumi K, Nagata K, Taketomi E, Ueyama K. Interobserver and intraobserver reliability of the Japanese Orthopaedic Association scoring system for evaluation of cervical compression myelopathy. Spine (Phila Pa 1976) 2001; 26: Ryan AG, Morrissey BM, Newcombe RG, Halpin SFS, Hourihan MD. Are T1 weighted images helpful in MRI of cervical radiculopathy? Br J Radiol 2004; 77: Gilbert JW, Wheelera GR, Lingreena RA, Johnsona RK, Scheinera SJ, Gibbsa RD. Upright weight-bearing cervical flexion/extension dynamic magnetic resonance imaging: case report and review of the literature. Eur J Radiol Extra 2006; 60: Elsig JPJ, Kaech DL. Dynamic imaging of the spine with an open upright MRI: present results and future perspectives of fmri. Eur J Orthop Surg Traumatol 2007; 17: W140 AJR:197, July 2011