Intraobserver and interobserver agreement of the interpretation of pediatric chest radiographs

The objective of this study is to quantify the magnitude of intraobserver and interobserver agreement among physicians for the interpretation of pneumonia on pediatric chest radiographs. Chest radiographs that produced discordant interpretations between the emergency physician and the radiologist's final interpretation were identified for patients aged 1–4 years. From 24 radiographs, eight were randomly selected as study radiographs, and 16 were diversion films. Study participants included two pediatric radiologists, two senior emergency medicine physicians, and two junior fellowship-trained pediatric emergency medicine physicians. Each test included 12 radiographs: the eight study radiographs and four randomly interspersed diversion radiographs, and each radiograph was paired with a written clinical vignette. Testing was repeated on four occasions, separated by ≥2 weeks. The dependent variable was the interpretation of presence or absence of pneumonia; primary analysis done with Cohen's kappa (95% confidence intervals). Intraobserver agreement was good for pediatric radiologists (kappa = 0.87; 95% CI 0.60–0.99) for both but was lower for senior emergency physicians (mean kappa = 0.68; 95% CI 0.40–0.95) and junior pediatric emergency physicians (mean kappa = 0.62; 95% CI 0.35–0.98). Interobserver agreement was fair to moderate overall; between pediatric radiologists, kappa = 0.51 (0.39–0.64); between senior emergency physicians, kappa = 0.55 (0.41–69), and between junior pediatric emergency medicine physicians, kappa = 0.37 (0.25–0.51). Practicing emergency clinicians demonstrate considerable intraobserver and interobserver variability in the interpretation of pneumonia on pediatric chest radiographs.     

Power Doppler ultrasound of rheumatoid synovitis: quantification of vascular signal and analysis of interobserver variability

Objective  The objective of this study was to assess interobserver uncertainties in power Doppler (PD) examination of the fingers of patients with rheumatoid arthritis (RA), by separating the source of the discrepancy into (1) acquisition of the images and (2) criteria for assessment of the images. Materials and methods  Twenty patients who had been diagnosed with RA were enrolled in this study. Ultrasound examinations were performed by one inexperienced and two experienced sonographers. Interobserver variation was measured using a conventional semiquantitative image grading scale. Interobserver variation of the quantitative PD (QPD) index (the summation of the colored pixels in a region of interest) was also assessed. Results  The agreement was higher between the two experienced sonographers (kappa value of 0.8) than between experienced and inexperienced sonographers (kappa value, 0.6–0.7) in the semiquantitative image grading scale. Results suggest that the difference in the assessment on the image grading scale was due more to the difference in the acquisition of the images than to variations in the grading criteria between sonographers. An excellent relationship was noted between the image grading scale and the QPD index for Doppler signal with a Spearman’s coefficient of rank correlation of 0.83 (P < 0.0001). Conclusions  Interobserver discrepancies in the image grading and QPD index methods were due more to the difference in the acquisition of the image than to the grading criteria used. The QPD index seems to be as reliable as the image grading scale with reasonable interobserver agreement between experienced sonographers.     

Magnetic resonance imaging for evaluation of disease activity in Crohn’s disease: a systematic review

To systematically review the evidence on the accuracy of MRI for grading disease activity in Crohn’s disease (CD). The MEDLINE, EMBASE, CINAHL and Cochrane databases were searched for studies on the accuracy of MRI in grading CD compared to a predefined reference standard. Two independent observers scored all relevant data. Three disease stages were defined: remission, mild and frank disease. The accuracy rates of MRI per disease stage were calculated by means of a random-effects model. Seven studies were included from a search resulting in 253 articles. In total 140 patients (16 patients in remission, 29 with mild disease and 95 with frank disease) were used for data analysis. MRI correctly graded 91% (95% CI: 84–96%) of patients with frank disease, 62% (95% CI: 44–79) of patients with mild disease and 62% (95% CI: 38–84) of patients in remission. MRI more often overstaged than understaged disease activity; MRI overstaged disease activity in 38% of patients in remission, mostly as mild disease. Overstaging of mild disease was observed in 21%, understaging in 17%. MRI correctly grades disease activity in a large proportion of patients with frank disease. For patients in remission or with mild disease, MRI correctly stages disease activity in many patients (62%).     

Regional myocardial wall thickening and global ejection fraction in patients with low angiographic left ventricular ejection fraction assessed by visual and quantitative resting ECG-gated 99mTc-tetrofosmin single-photon emission tomography and magnetic resonance imaging

We investigated the use of visual and quantitative technetium 99m tetrofosmin ECG-gated single-photon emission tomography (SPET) for the assessment of regional myocardial wall thickening (WT) and left ventricular (LV) ejection fraction (EF) in comparison with gated magnetic resonance imaging (MRI) in patients with a low angiographic LVEF. Gated SPET using ^99mTc-labelled flow tracers offers potential for simultaneous assessment of myocardial perfusion and LV function. Few data are available on the use of visual and quantitative gated SPET in patients with low LVEF. In this study 21 patients with low angiographic LVEF (mean 37%±5%) were studied. Resting gated ^99mTc-tetrofosmin SPET and gated MRI were performed within 48 h. WT was assessed by visual interpretation (five point score) and quantitative analysis based on count increase. There was good agreement for EF measurements by MRI and gated SPET (mean EF: 33%±12% vs 35%±11%, r = 0.86, P<0.001). Areas under receiver operator characteristic curves (AUC) for differentiation between MRI WT score points ranged from 0.60 to 0.66 for visual SPET WT analysis, from 0.59 to 0.71 for delta count increase values and from 0.46 to 0.60 for % WT, indicating substantial overlap between WT categories. Absolute agreement for visual WT between MRI and gated SPET ranged from 25% to 57% (kappa 0.03–0.25) depending on tracer uptake, and was limited in areas with moderate to severe perfusion defects (kappa 0.03–0.13). It is concluded that gated SPET provided reliable estimates of regional WT and global function in patients with low angiographic LVEF. Received 15 October 1997 and in revised form 31 January 1998     

Prospective comparison of helical CT with angiography in pulmonary embolism: global and selective vascular territory analysis. Interobserver agreement

The objective of this prospective study was to evaluate the sensitivity, specificity, positive and negative predictive values, and interobserver agreement in the diagnosis of pulmonary embolism with helical CT, compared with pulmonary angiography, for both global results and for selective vascular territories. Helical CT and pulmonary angiography were performed on 66 consecutive patients with clinical suspicion of pulmonary embolism. The exams were blindly interpreted by a vascular radiologist and by two independent thoracic radiologists. Results were analyzed for the final diagnosis as well as separately for 20 different arterial territories in each patient. Pulmonary angiography revealed embolism in 25 patients (38%); 48% were main, 28% lobar, 16% segmental, and 8% subsegmental. The sensitivity, specificity, and positive and negative predictive values of helical CT for observer 1 were, respectively, 91, 81.5, 75, and 94%; in 7.5% of the patients the exam was considered indeterminate. For observer 2 the values were, respectively, 88, 86, 81.5, and 91%; in 9% of the patients the exam was considered indeterminate. Main arteries were considered as non-valuable in 0–0.8%, the lobar in 1.5%, the segmental in 7.5–8.5%, and the subsegmental in 55–60%. Interobserver agreement for the final diagnosis was 80% (kappa 0.65). For each vascular territory, this was 98% (kappa 0.91) for main arteries, 92% (kappa 0.78) for lobar arteries, 79% (kappa 0.56) for segmental arteries, and 59% (kappa 0.21) for subsegmental arteries. Helical CT is a reliable method for pulmonary embolism diagnosis, with good interobserver agreement for main, lobar, and segmental territories. Worse results are found for subsegmental arteries, with high incidence of non-valuable branches and poor interobserver agreement. Electronic Publication     

Chronic post-traumatic osteomyelitis of the lower extremity: comparison of magnetic resonance imaging and combined bone scintigraphy/immunoscintigraphy with radiolabelled monoclonal antigranulocyte antibodies

Objective. A retrospective study of the validity of combined bone scintigraphy (BS) and immunoscintigraphy (IS) using ^99mTc-labelled murine antigranulocyte antibodies (MAB) and magnetic resonance imaging (MRI) in chronic post- traumatic osteomyelitis. Design and patients. The results of MRI and combined BS/IS of 19 lesions in 18 patients (13 men, 5 women; mean age 45 years, range 27–65 years) were independently evaluated by two radiologists and one nuclear medicine physician with regard to bone infection activity and extent. The patient group was a highly selective collection of clinical cases: the average number of operations conducted because of relapsing infection was eight (range 2–27), the average time interval between the last surgical intervention and the present study was 6.5 years (range 3 months to 39 years), and from the first operation was 14 years (range 1.5–42 years). Interobserver agreement on MRI was measured by kappa statistics. Sensitivity, specificity, accuracy, positive predictive value (PPV) and negative predictive value (NPV) were calculated for MRI and the nuclear medicine studies. Results. For MRI/nuclear medicine, a sensitivity of 100%/77%, a specificity of 60%/50%, an accuracy of 79%/61%, a PPV of 69%/58% and a NPV of 100%/71% were calculated. Four MR examinations were false positives because of postsurgical granulation tissue. A high degree of interobserver agreement was found on MRI (κ=0.88). A low-grade infection was missed on two scintigrams, while four were false positive because of ectopic haematopoietic bone marrow, and in one examination the anatomical distortion resulted in an inaccurate assignment of the uptake leading to false positive findings. Image analysis was frequently hindered by susceptibility artefacts due to residual abrasions of metallic implants after removal of orthopaedic devices (15/18 patients); this led to limited assessment in 17% (3/18 patients). Conclusion. Acute activity in a chronic osteomyelitis can be excluded with high probability if the MRI findings are negative. In the first postoperative year fibrovascular scar cannot be distinguished accurately from reactivated infection on MRI and scintigraphy may improve the accuracy of diagnosis. MRI is more sensitive in low-grade infection during the later course than combined BS/IS. Scintigraphic errors due to ectopic, peripheral, haematopoietic bone marrow can be corrected by MRI. Received: 17 December 1999 Revision requested: 25 January 2000 Revision received: 15 March 2000 Accepted: 4 April 2000     

Image interpretation in CT of laryngeal carcinoma: a study on intra- and interobserver reproducibility

The object of our study was to evaluate the intra- and interobserver reproducibility of the interpretation of CT examinations of laryngeal carcinoma. The CT examinations of 100 laryngeal cancers were retrospectively reviewed twice by two independent reviewers. Involvement of different structures was assessed, using a standard scoring form. Statistical analysis was done using the Wilcoxon signed rank test and Cohen's kappa. A borderline significant difference between the observers (p < 0.04) was present. Fair to substantial intraobserver reproducibility (kappa = 0.29–0.86), and fair to substantial interobserver reproducibility (average kappa = 0.26–0.74) were found for most laryngeal structures when a dichotomous categorical scale was used. On the average somewhat lower, but still fair to substantial, intraobserver (kappa = 0.36–0.72), and fair to moderate interobserver (average kappa = 0.29–0.47) reproducibility, were found when a nominal or ordinal categorical scale was used. In conclusion, the interpretation of CT images of laryngeal tumors is reproducible. Received 4 November 1996; Revision received 27 December 1996; Accepted: 2 January 1997     

Inadequacy of manual measurements compared to automated CT volumetry in assessment of treatment response of pulmonary metastases using RECIST criteria

The purpose of this study was to compare relative values of manual unidimensional measurements (MD) and automated volumetry (AV) for longitudinal treatment response assessment in patients with pulmonary metastases. Fifty consecutive patients with pulmonary metastases and repeat chest multidetector-row CT (median interval=2 months) were independently assessed by two radiologists for treatment response using Response Evaluation Criteria In Solid Tumours (RECIST). Statistics included relative measurement errors (RME), intra-/interobserver correlations, limits of agreement (95% LoA), and kappa. A total of 202 metastases (median volume=182.22 mm³; range=3.16–5,195.13 mm³) were evaluated. RMEs were significantly higher for MD than for AV (intraobserver RME=2.34–3.73% and 0.15–0.22% for MD and AV respectively; P<0.05. Interobserver RME=3.53–3.76% and 0.22–0.29% for MD and AV respectively; P<0.05). Overall correlation was significantly better for AV than for MD (P<0.05). Intraobserver 95% LoAs were −1.85 to 1.75 mm for MD and −11.28 to 9.84 mm³ for AV. The interobserver 95% LoA were −1.46 to 1.92 mm for MD and −11.17 to 9.33 mm³ for AV. There was total intra-/interobserver agreement on response using AV (κ=1). MD intra- and interobserver agreements were 0.73–0.84 and 0.77–0.80 respectively. Of the 200 MD response ratings, 28 (14/50 patients) were discordant. Agreement using MD dropped significantly from total remission to progressive disease (P<0.05). We therefore conclude that AV allows for better reproducibility of response evaluation in pulmonary metastases and should be preferred to MD in these patients.     

Interobserver Variability in the Evaluation of Chronic Mesenteric Ischemia with Gadolinium-enhanced MR Angiography

RATIONALE AND OBJECTIVES: The purpose of this study was to assess interobserver variability in the interpretation of gadolinium-enhanced magnetic resonance (MR) angiograms of splanchnic vessels in patients suspected of having chronic mesenteric ischemia (CMI). MATERIALS AND METHODS: Two readers blinded to the initial interpretation retrospectively reviewed gadolinium-enhanced MR angiograms obtained for suspected CMI in 26 patients (20 women and six men; age range, 23-77 years; mean age, 61 years) who also underwent conventional angiography. Each reader graded the degree of stenosis based on the percentage diameter reduction of the celiac artery (CA), superior mesenteric artery (SMA), and inferior mesenteric artery (IMA) by using a five-point ordinal scale: 0, no stenosis: 1, mild stenosis (<50%); 2, moderate stenosis (50%-75%); 3, severe stenosis (>75%); 4, occluded artery. Using the conventional angiogram as a reference standard, authors determined sensitivity and specificity for each observer, assigning two thresholds (grades 2 and 3) as significant stenoses. A kappa statistic (kappa) measured interobserver agreement. RESULTS: With grade 2 stenosis used as a threshold, cumulative accuracies for detecting significant stenosis were 0.95 (95% confidence interval, 0.86-0.99) for reader A and 0.97 (0.88-1.0) for reader B. Interobserver agreement for grading proximal splanchnic stenosis was 0.90 for CA, 0.92 for SMA, and 0.48 for IMA. CONCLUSION: Gadolinium-enhanced MR angiography is reproducibly accurate for detection of proximal splanchnic artery stenosis, with good to excellent interobserver agreement.     

Diffuse optical tomography of the breast: preliminary findings of a new prototype and comparison with magnetic resonance imaging

This paper presents an evaluation of a prototype diffuse optical tomography (DOT) system. Seventeen women with 18 breast lesions (10 invasive carcinomas, 2 fibroadenomas, and 6 benign cysts; diameters 13–54 mm) were evaluated with DOT and magnetic resonance imaging (MRI). A substantial fraction of the original 36 recruited patients could not be examined using this prototype due to technical problems. A region of interest (ROI) was drawn at the lesion position as derived from MRI and at the mirror image site in the contralateral healthy breast. ROIs were assessed quantitatively and qualitatively by two observers independently in two separate readings. Intra- and interobserver agreements were calculated using kappa statistics (k) and intraclass correlation coefficients (ICCs). Discriminatory values for presence of malignancy were determined by receiver operating characteristic (ROC) analyses. Intraobserver agreements were excellent (k 0.88 and 0.88; ICC 0.978 and 0.987), interobserver agreements were good to excellent (k 0.77–0.95; ICC 0.96–0.98). Discriminatory values for presence of malignancy were 0.92–0.93 and 0.97–0.99 for quantitative and qualitative ROC analysis, respectively. This DOT system has the potential to discriminate malignant from benign breast tissue in a reproducible qualitative and quantitative manner. Important technical improvements are required before this technique is ready for clinical application.     

MRI of the alar and transverse ligaments in whiplash-associated disorders (WAD) grades 1–2: high-signal changes by age,gender, event and time since trauma

Introduction  This study describes the prevalence of high-signal changes at magnetic resonance imaging (MRI) of the alar and transverse ligaments in whiplash-associated disorders (WAD) grades 1–2 in relation to age, gender, spinal degeneration, type of trauma event and time since trauma. Materials and methods  In 1,266 consecutive WAD1–2 patients (779 women, 487 men; mean age 42 years) referred from clinicians, high-signal changes in the alar and transverse ligaments at high-resolution proton-weighted MRI were prospectively graded 0–3 based on a previously reported, reliable grading system. Type of event according to The International Statistical Classification of Diseases and Related Health Problems and time of trauma were obtained from referral letters. Results  MRI showed grades 2–3 alar ligament changes in 449 (35.5%; 95% confidence interval (CI), 32.8 to 38.1%) and grades 2–3 transverse ligament changes in 311 (24.6%; 95% CI, 22.2% to 26.9%) of the 1,266 patients. Grades 2–3 changes were more common in men than women, odds ratio 1.9 (95% CI, 1.5 to 2.5) for alar and 1.5 (95% CI, 1.1 to 2.0) for transverse ligament changes. High-signal changes were not related to age, spinal degeneration, type of trauma event or time since trauma (median 5 years). Unilateral changes were more often left- than right-sided. Conclusions  High-signal changes of the alar and transverse ligaments are common in WAD1-2 and unlikely to represent age-dependent degeneration. Their male and left-side preponderance cannot be explained by variation in ligament stretching or image artefacts. Further studies are needed to clarify whether such changes are caused by trauma.     

Visual MRI Grading System to Evaluate Atrophy of the Supraspinatus Muscle

Objective

To investigate the interobserver reproducibility and diagnostic feasibility of a visual grading system for assessing atrophy of the supraspinatus muscle on magnetic resonance imaging (MRI).

Materials and Methods

Three independent radiologists retrospectively evaluated the occupying ratio of the supraspinatus muscle in the supraspinatus fossa on 192 shoulder MRI examinations in 188 patients using a 3-point visual grading system (1, ≥ 60%; 2, 30-59%; 3, < 30%) on oblique sagittal T1-weighted images. The inter-reader agreement and the agreement with the reference standard (3-point grades according to absolute occupying ratio values quantitatively measured by directly contouring the muscles on MRI) were analyzed using weighted kappa. The visual grading was applied by a single reader to a group of 100 consecutive patients who had undergone rotator cuff repair to retrospectively determine the association between the visual grades at preoperative state and postsurgical occurrences of retear.

Results

The inter-reader weighted kappa value for the visual grading was 0.74 when averaged across three reader pairs (0.70-0.77 for individual reader pairs). The weighted kappa value between the visual grading and the reference standard ranged from 0.75 to 0.83. There was a significant difference in retear rates of the rotator cuff between the 3 visual grades of supraspinatus muscle atrophy on MRI in univariable analysis (p < 0.001), but not in multivariable analysis (p = 0.026).

Conclusion

The 3-point visual grading system may be a feasible method to assess the severity of supraspinatus muscle atrophy on MRI and assist in the clinical management of patients with rotator cuff tear.     

MRI screening for acoustic neuroma without gadolinium: value of 3DFT-CISS sequence

To evaluate the efficacy of a gradient-echo sequence (3DFT-CISS) in the diagnosis of acoustic neuromas, two independent observers twice reviewed the images of the temporal bones of 83 patients. Contrast-enhanced T1-weighted spin echo images were used as the reference, showing 18 acoustic neuromas, including 5 purely intracanalicular and one intralabyrinthine tumours. High sensitivity (89-94 %), specificity (94–97 %) and accuracy (94–95 %) were found. Intraobserver (kappa 0.93–1) and interobserver (kappa 0.83–0.84) reproducibility were very good. The smallest intracanalicular tumour was overlooked twice by both observers; the intralabyrinthine tumour once by one observer. All tumours were detected with a less stringent decision criterion, at the expense of lower specificity. Received: 2 September 1996 Accepted: 19 November 1996     

Fluid-fluid levels in bone neoplasms: variation of T1-weighted signal intensity of the superior to inferior layers—diagnostic significance on magnetic resonance imaging

The diagnostic relevance of the relative T1-weighted (T1W) and T2-weighted (T2W)/short tau inversion recovery (STIR) MRI signal intensity characteristics of the superior to inferior fluid layers within fluid-fluid levels (FFLs) found in bone tumours was investigated. A retrospective analysis was performed of MRI studies of 2,568 patients presenting with a suspected bone tumour over an 8-year period. Final diagnosis was made by biopsy/surgical resection or characteristic imaging/clinical findings. Subjects were divided by the absence/presence of FFLs and benign/malignant histology. Cases with FFLs were sub-categorised by the relative signal intensity of the superior/inferior layer as high/low or low/high on T1W and T2W/STIR sequences. Out of the total of 2,568 cases, 214 (8.3%, CI 7.3–9.5%) had FFLs and 2,354 (91.7%, CI 90.5–92.7%) had no FFLs. All 214 cases with FFLs had T2W/STIR sequences available, all demonstrating high/low signal intensity characteristics; 135/214 (63.1%, CI 56.2–69.6%) were benign and 79/214 (36.9%, CI 30.4–43.8%) were malignant. Out of the 214 patients, 151 had T1W sequences performed; 52 showed high/low signal intensity, of which 30 (57.7%, CI 34.2–71.3%) were benign and 22 (42.3%, CI 28.7–56.8%) were malignant (P = 0.06 compared with no FFL group); 50 showed low/high signal intensity, of which 40 (80%, CI 66.3–90.0%) were benign and ten (20%, CI 10.0–33.7%) were malignant (P = 0.0000, compared with the no FFL group). The low/high and high/low groups had a significantly greater proportion of benign and malignant lesions, respectively (P = 0.015). In conclusion, all FFLs showed high/low signal intensity characteristics on T2W/STIR sequences. Low/high signal on T1W was significantly associated with benign disease. Malignancy may occur slightly more frequently with high/low signal on T1W.     

MRI of non-specific low back pain and/or lumbar radiculopathy: do we need T1 when using a sagittal T2-weighted Dixon sequence?

To show that for the MRI workup of non-specific low back pain and/or lumbar radiculopathy, the acquisition of T1-weighted sequences in the sagittal plane could be waived when using an FSE T2-weighted Dixon sequence. Three musculoskeletal radiologists retrospectively reviewed fifty lumbar spine MRI examinations performed for non-specific low back pain and/or lumbar radiculopathy. Two protocols were separately analyzed in the sagittal plane: a standard protocol (T1-weighted, in-phase, and water-only images of an FSE T2-weighted Dixon sequence) and a simplified protocol (fat-only, in-phase, and water-only images of an FSE T2-weighted Dixon sequence). Eight items usually assessed on T1-weighted sequences were analyzed for each of the vertebrae (n = 250), vertebral endplates (n = 500), vertebral corners (n = 1000), foramina (n = 500), lamina (n = 500), and facet joints (n = 500). Interchangeability of these protocols was tested using the individual equivalence index. A decrease in interobserver agreement of ≥ 5% when one reader used the simplified protocol compared with when both readers used the standard protocol was considered clinically significant. Interreader and intrareader agreement were assessed using kappa statistics. Rates of findings with each protocol were compared using odd ratios. The standard and simplified protocols were interchangeable (range of upper bound of the 95%CI of individual equivalence index = 0.25 to 1.38%). Intraprotocol and interprotocol interreader kappa values were similar (0.253–0.671 vs. 0.236–0.723, respectively). Rates of findings were not statistically significantly different (p ≥ 0.074), or were higher with the simplified protocol (p ≤ 0.036). In our target population, a single sagittal T2-weighted Dixon sequence may replace the recommended combination of T1-, T2-, and fat-suppressed T2-weighted sequences. • In patients with non-specific low back pain or lumbar radiculopathy, spine MRI in the sagittal plane could be limited to a single FSE T2-weighted Dixon sequence, hereby reducing the acquisition time. • A simplified protocol of spine MRI in the sagittal plane combining FSE T2-weighted Dixon sequence provides the same information as a standard protocol including T1-, T2-, and fat-suppressed T2-weighted sequences for the workup of degenerative lumbar spine lesions. • For some findings shown on the simplified protocol, such as focal bone marrow replacement lesions or signs of infection, additional sequences including pre- and post-contrast T1-weighted sequences may be required, as is currently the case when using the standard protocol.     

MR Reproducibility in the Assessment of Uterine Fibroids for Patients Scheduled for Uterine Artery Embolization

Magnetic resonance imaging (MRI) is increasingly applied in the evaluation of uterine fibroids. However, little is known about the reproducibility of MRI in the assessment of uterine fibroids. This study evaluates the inter- and intraobserver variation in the assessment of the uterine fibroids and concomitant adenomyosis in women scheduled for uterine artery embolization (UAE). Forty patients (mean age: 44.5 years) with symptomatic uterine fibroids who were scheduled for UAE underwent T₁- and T₂-weighted MRI. To study inter- and intraobserver agreement 40 MR images were evaluated independently by two observers and reevaluated by both observers 4 months later. Inter- and intraobserver agreement was calculated using Cohen’s κ statistic and intraclass correlation coefficient for categorical and continuous variables, respectively. Inter-observer agreement for uterine volumes (κ = 0.99, p < 0.0001), dominant fibroid volumes (κ = 0.98, p ≤ 0.0001), and number of fibroids (κ = 0.88; CI, 0.77–0.93; p < 0.0001) was excellent. For the T₁- and T₂-weighted signal intensity of the dominant fibroid there was good agreement between the observers (87%; 95% CI, 71.9%–95.6%) and the intraobserver agreement was good for observer A (95%; 95% CI, 83.1%–99.4%) and moderate for observer B (κ = 0.47). The interobserver agreement with respect to the presence of adenomyosis was good (κ = 0.73, p < 0.0001), while both intraobserver agreements were fair to moderate (observer A, κ = 0.55, p = 0.0003; and observer B, κ = 0.66, p < 0.0001). In conclusion, MRI criteria used for the selection of suitable UAE patients show good inter- and intraobserver reproducibility.     

Interobserver agreement in the assessment of cerebral atrophy on CT using bicaudate and sylvian-fissure ratios

To assess the interobserver variability of cerebral-atrophy measures on CT, three investigators measured the bicaudate ratio (BCR) and the sylvian-fissure ratio (SFR) on 20 CT studies of patients with ischaemic stroke. The intraclass correlation coefficient of BCR measurements was 0.82 [95 % confidence interval (CI) 0.75–0.94], and that of SFR measurements 0.69 (95 % CI 0.57–0.89). The range of pairwise-calculated Pearson correlation coefficients was smaller for measurement of the BCR (0.89–0.92) than for the SFR measurements (0.66–0.84). Received: 2 June 1998 Accepted: 4 October 1998     

The patellotrochlear index: a new index for assessing patellar height

The radiological methods to determine patellar height described in the literature are variable, not reliable and depend on the chosen ratio. The purpose of this paper is to describe another method of measuring patellar height on sagittal MRI using the true articular cartilage patellotrochlear relationship. An analysis of magnetic resonance (MR) examinations of 66 consecutive patients was performed. The most common diagnoses were meniscal or anterior cruciate ligament pathologies. No patient suffered from patellofemoral complaints. Measurements on sagittal MR images included different parameters using the articular cartilage of the patella and the trochlea. The ratio patella : trochlea of the cartilage baselines was measured in percentages and described as patellotrochlear index. The measurements were assessed at two different times by three raters under blinded conditions. The mean patellotrochlear index was 31.7% (CI: 12.5–50.0; range −5.0 to 61.1%; SD ±11.6). The intraobserver variability showed only in the “second observer” a difference of the mean values of the two different measurements (t=2.189; P=0.032). The interobserver correlation was high and significant (0.663–0.893; P=0.000). Our results indicate that the patellotrochlear index is a reliable and precise method to determine the exact articular correlation of the patellofemoral joint and the patellar height. The results represent the average patellotrochlear index in the normal population without patellofemoral complaints. Measurements of the articular cartilage congruence can be helpful to define an underlying pathology of patellar height, such as patella alta or infera.     

Observer variation in MRI evaluation of patients suspected of lumbar disk herniation

OBJECTIVE: Our objective was to assess observer variation in MRI evaluation in patients suspected of lumbar disk herniation. SUBJECTS AND METHODS: Two experienced neuroradiologists independently evaluated 59 consecutive patients with lumbosacral radicular pain. Per patient, three levels (L3-L4 through L5-S1) and the accompanying roots were evaluated on both sides. For each segment, the presence of a bulging disk or a herniation and compression of the root was reported. Images were interpreted twice: once before and once after disclosure of clinical information. Interobserver agreement was expressed as unweighted kappa values. RESULTS: Without clinical information, interobserver agreement for the presence of herniation or bulging disk was moderate (full agreement, 84%; kappa = 0.63; 95% confidence interval [CI], 0.53-0.72). Of a total of 352 segments evaluated, there was disagreement on 58 segments (17%): bulging disk versus no defect in 26 (7.4%), bulging disk versus herniation in five (1.4%), and hernia versus no defect in 27 (7.7%). With clinical information, twice as many bulging disks were reported but no new herniations were detected. Agreement slightly decreased, but not significantly (full agreement, 77%; kappa = 0.59; 95% CI, 0.49-0.69; p = 0.12). CONCLUSION: On average, more than 50% of interobserver variation in MRI evaluation of patients with lumbosacral radicular pain is caused by disagreement on bulging disks. Knowledge of clinical information does not influence the detection of herniations but lowers the threshold for reporting bulging disks.     

Fat-suppressed 3D spoiled gradient-echo MRI and MDCT arthrography of articular cartilage in patients with hip dysplasia

OBJECTIVE: Our objective was to assess the diagnostic ability of MDCT arthrography for acetabular and femoral cartilage lesions in patients with hip dysplasia. MATERIALS AND METHODS: A disorder of the articular cartilage was evaluated in 20 hips of 18 patients with acetabular dysplasia who did not have osteoarthritis or who had early stage osteoarthritis before undergoing pelvic osteotomy surgery. The findings on fat-suppressed 3D fast spoiled gradient-echo MRI and MDCT arthrography of the hip were evaluated by two independent observers, and sensitivity, specificity, and accuracy were determined using arthroscopic findings as the standard of reference. Kappa values were calculated to quantify the level of interobserver agreement. RESULTS: The sensitivity and specificity for the detection of any cartilage disorder (grade 1 or higher) were (observer 1/observer 2) 49%/67% and 89%/76%, respectively, on MRI, and 67%/67% and 89%/82%, respectively, on CT arthrography. The sensitivity and specificity for the detection of cartilage lesions with substance loss (grade 2 or higher) were (observer 1/observer 2) 47%/53% and 92%/87%, respectively, on MRI, and 70%/79% and 93%/94%, respectively, on CT arthrography. CT arthrography provided significantly higher sensitivity in the detection of grade 2 or higher lesions than MRI for both observers. Interobserver agreement in the detection of grade 2 or higher cartilage lesions was moderate (kappa = 0.53) on MRI and substantial (kappa = 0.78) on CT. CONCLUSION: MDCT arthrography is a sensitive and reproducible method for assessing articular cartilage lesions with substance loss in patients with hip dysplasia.