Yield of urinary tract cancer diagnosis with repeat CT urography in patients with hematuria.

OBJECTIVE. The purpose of this study was to evaluate the yield of repeat CT urography (CTU) in detecting urinary tract malignancies in patients with hematuria. MATERIALS AND METHODS. A review of 5525 patients who underwent CTU between 2000 and 2011 revealed 751 (13.6%) patients who underwent repeat CTU. We excluded 127 patients with more than 3 years between examinations, 409 with nonhematuria indications, and 13 with less than 1 year of follow-up from a negative repeat examination. An additional 54 patients with malignancy diagnosed on the initial evaluation were excluded, leaving 148 patients in the study cohort (77 men and 71 women; mean age, 57 years). Patients were categorized on the basis of the presence or absence of findings suspicious for malignancy on initial CTU reports. Repeat CTU reports were correlated with cystoscopy, pathology, and clinical follow-up to determine the incidence of malignancy. Examinations negative for malignancy were confirmed with at least 1 year of clinical follow-up. CTU examinations of patients diagnosed with malignancy on repeat examination were reviewed by two radiologists in consensus. RESULTS. Initial CTU showed no findings suspicious for malignancy in 103 (70%) of 148 patients; of these, none had malignancy identified on repeat CTU. Among 45 (30%) patients with suspicious initial CTU findings, four malignancies were identified on repeat CTU (8.9%). Three were incidental to the initial suspicious finding; in retrospect, two were present on the initial CTU examination. CONCLUSION. In patients with hematuria, repeat CTU within 3 years is unlikely to show a urinary tract malignancy. These results support currently published guidelines.

Impact of a Structured Report Template on the Quality of MRI Reports for Rectal Cancer Staging.

OBJECTIVE: The purpose of this study is to assess the impact of implementing a structured report template on the quality of MRI reports for rectal cancer staging. MATERIALS AND METHODS: After excluding examinations performed after surgery or neoadjuvant therapy, we analyzed all rectal cancer staging MRI reports finalized at an academic medical center 12 months before and after an intervention consisting of implementing a structured report template integrated into the institution's speech recognition system. The primary outcome measure was the quality of rectal cancer staging MRI reports classified as optimal, satisfactory, or unsatisfactory, on the basis of the documentation of 14 quality measures predefined by a consensus of the institution's abdominal radiology subspecialists. Chi-square and t tests were used to assess differences in report quality and documentation of each discrete quality measure before and after the intervention. RESULTS: The study cohort included 106 MRI reports from 104 patients (mean age, 60 years; 58.5% male); 52 (49.1%) of the reports were completed before implementation of the structured report template. After implementation, the proportion of total reports classified as optimal or satisfactory increased from 38.5% (20/52) to 70.4% (38/54) (p = 0.0010). No reports generated before the intervention were classified as optimal, whereas 40.7% (22/54) of reports were classified as optimal after the intervention. CONCLUSION: Implementation and voluntary use of a structured report template improved the quality of MRI reports for rectal cancer staging compared with free-text format.

Use of CT and MRI in emergency department patients with acute pancreatitis.

PURPOSE: Assess the utility of CT and MRI in patients with acute pancreatitis (AP) presenting to emergency department (ED). MATERIALS AND METHODS: In this Institutional Review Board-approved retrospective study, we identified all patients with AP from March 2012 through February 2013 in ED of a teaching hospital with approximately 60,000 annual visits. Patients were initially identified via ICD-9 code for AP (577.0); diagnosis was confirmed by chart review using established diagnostic criteria (presence of two of the following: typical abdominal pain, elevated lipase/amylase >3 times normal, or imaging findings of pancreatitis). Abdominal CT or MRI obtained in the ED and within 24 h of admission was reviewed by a fellowship-trained abdominal radiologist. RESULTS: Of 101 patients admitted with AP (60 women, 41 men; mean age 52 years, range 20-89), 63 (62.4%) underwent imaging; only one (1.6%) showed pancreatic necrosis. 88 (87.1%) patients could have been clinically diagnosed without imaging based on presence of abdominal pain and elevated laboratory values; 13 (12.9%) required imaging for diagnosis. Of 88 patients who met AP diagnostic criteria without imaging, 50 (56.8%) nonetheless underwent imaging, with AP without necrosis seen in 34 (68.0%), pancreatic necrosis in one (2.0%), sequelae of prior AP in four (8.0%), and no abnormality in 11 (22.0%). CONCLUSION: Early imaging is common in patients with AP, even when the diagnosis can be established based on non-imaging criteria, rarely demonstrating pancreatic necrosis. Reducing overuse of early imaging in patients with confident diagnosis of AP may improve quality and reduce waste.

Necrotizing pancreatitis: diagnosis, imaging, and intervention.

Acute necrotizing pancreatitis is a severe form of acute pancreatitis characterized by necrosis in and around the pancreas and is associated with high rates of morbidity and mortality. Although acute interstitial edematous pancreatitis is diagnosed primarily on the basis of signs, symptoms, and laboratory test findings, the diagnosis and severity assessment of acute necrotizing pancreatitis are based in large part on imaging findings. On the basis of the revised Atlanta classification system of 2012, necrotizing pancreatitis is subdivided anatomically into parenchymal, peripancreatic, and combined subtypes, and temporally into clinical early (within 1 week of onset) and late (>1 week after onset) phases. Associated collections are categorized as "acute necrotic" or "walled off" and can be sterile or infected. Imaging, primarily computed tomography and magnetic resonance imaging, plays an essential role in the diagnosis of necrotizing pancreatitis and the identification of complications, including infection, bowel and biliary obstruction, hemorrhage, pseudoaneurysm formation, and venous thrombosis. Imaging is also used to help triage patients and guide both temporizing and definitive management. A "step-up" method for the management of necrotizing pancreatitis that makes use of imaging-guided percutaneous catheter drainage of fluid collections prior to endoscopic or surgical necrosectomy has been shown to improve clinical outcomes. The authors present an algorithmic approach to the care of patients with necrotizing pancreatitis and review the use of imaging and interventional techniques in the diagnosis and management of this pathologic condition.

PET/CT-guided percutaneous liver mass biopsies and ablations: targeting accuracy of a single 20 s breath-hold PET acquisition.

AIM: To determine whether a single 20 s breath-hold positron-emission tomography (PET) acquisition obtained during combined PET/computed tomography (CT)-guided percutaneous liver biopsy or ablation procedures has the potential to target 2-[(18)F]-fluoro-2-deoxy-d-glucose (FDG)-avid liver masses as accurately as up to 180 s breath-hold PET acquisitions. MATERIALS AND METHODS: This retrospective study included 10 adult patients with 13 liver masses who underwent FDG PET/CT-guided percutaneous biopsies (n = 5) or ablations (n = 5). PET was acquired as nine sequential 20 s, monitored, same-level breath-hold frames and CT was acquired in one monitored breath-hold. Twenty, 40, 60, and 180 s PET datasets were reconstructed. Two blinded readers marked tumour centres on randomized PET and CT datasets. Three-dimensional spatial localization differences between PET datasets and either 180 s PET or CT were analysed using multiple regression analyses. Statistical tests were two-sided and p < 0.05 was considered significant. RESULTS: Targeting differences between 20 s PET and 180 s PET ranged from 0.7-20.3 mm (mean 5.3 ± 4.4 mm; median 4.3) and were not statistically different from 40 or 60 s PET (p = 0.74 and 0.91, respectively). Targeting differences between 20 s PET and CT ranged from 1.4-36 mm (mean 9.6 ± 7.1 mm; median 8.2 mm) and were not statistically different from 40, 60, or 180 s PET (p = 0.84, 0.77, and 0.35, respectively). CONCLUSION: Single 20 s breath-hold PET acquisitions from PET/CT-guided percutaneous liver procedures have the potential to target FDG-avid liver masses with equivalent accuracy to 180 s summed, breath-hold PET acquisitions and may facilitate strategies that improve image registration and shorten procedure times.

Detection and localization of bile duct leaks after cholecystectomy using Gd-EOB-DTPA-enhanced MR cholangiography: retrospective study of 16 patients.

OBJECTIVE: To determine if gadolinium-ethoxybenzyl-diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance (MR) cholangiography can detect and localize bile duct leaks in postcholecystectomy patients. MATERIAL AND METHODS: Four blinded independent radiologists performed a retrospective review of 16 consecutive patients who underwent MR cholangiography with intravenous Gd-EOB-DTPA for the evaluation of possible biliary leak. Image quality, ductal opacification, and presence and location of any bile leak were evaluated. An independent observer determined the criterion standard using a consensus of all chart, clinical, and imaging findings. RESULTS: All 6 bile leaks confirmed at endoscopic retrograde cholangiopancreatography were diagnosed by all reviewers (sensitivity, 100%). Of the 10 patients with no leak, only one reader incorrectly diagnosed a bile leak in a single case (specificity, 98%). The accuracy for detection of the site of leak with Gd-EOB-DTPA-enhanced MR cholangiography was 80%. CONCLUSION: Gadolinium-EOB-DTPA-enhanced MR can detect bile leaks with a high sensitivity and specificity.

Imaging female infertility.

The purpose of this pictorial review is to discuss causes of female infertility, in particular, those etiologies in which imaging plays a key role in detection. Included are disorders of cervical, ovarian, fallopian tube, and uterine origin. We also discuss the role of various imaging modalities including hysterosalpingography, pelvic ultrasonography, hysterosonography, and pelvic MR imaging in elucidating the cause of female infertility. Radiologists need to know the conditions to be aware of when these patients are sent for diagnostic imaging, as well as how to direct further management, if necessary, should an abnormality be detected.

Virtual unenhanced CT images acquired from dual-energy CT urography: accuracy of attenuation values and variation with contrast material phase.

AIM: To determine how representative virtual unenhanced (VNE) images are of true unenhanced (TNE) images when performing computed tomography (CT) urography on a dual-energy CT (DECT) system, and whether the images are affected by the contrast material phase. MATERIALS AND METHODS: In this retrospective, institutional review board-approved, Health Insurance Portability and Accountability Act (HIPAA)-compliant study, TNE were compared with VNE images derived from the nephrographic (VNEn) and excretory (VNEe) phases in 100 consecutive CT urograms. Two readers in consensus measured attenuation values of abdominal organs, fat, and renal lesions (>1 cm). Image noise was correlated with patient thickness. Detectability of renal stones was evaluated. Image quality and acceptability was assessed using a five-point scale. Expected dose saving by removing the TNE phase was calculated. RESULTS: VNE attenuation values of liver, renal parenchyma, and aorta were significantly different to TNE values (p < 0.05); spleen and fat attenuation values showed no significant difference. No significant difference was found between VNEn and VNEe images. Image noise was significantly greater in TNE images (p < 0.0001) and correlated with patient thickness. VNEn and VNEe images had sensitivities of 76.6 and 65.6% for detection of stones, identifying all stones greater than 3 and 4 mm, respectively. Both VNE images received significantly lower image quality scores than TNE images (p < 0.0001); however, the majority of images were deemed acceptable. The mean theoretical dose saving by removing the TNE phase was 35%. CONCLUSION: Although VNE images demonstrate high reader acceptability, accuracy of attenuation values and detection of small stones is limited. The contrast material phase, however, does not affect attenuation values. Further validation of VNE images is recommended prior to clinical implementation.

Multidetector-row computed tomography enterographic assessment of the ileal-anal pouch: descriptive radiologic analysis with endoscopic and pathologic correlation.

OBJECTIVE: To describe the multidetector-row computed tomography enterographic (MD-CTE) features of the ileal-anal pouch after ileal pouch anal anastomosis (IPAA) surgery and correlate them with pouch endoscopy and histopathologic findings. METHODS: All MD-CTE examinations performed on patients who underwent IPAA from July 1, 2005 to December 1, 2010 (n = 35; 16 [45.7%] men; mean age, 37.7 years; age range, 22-72 years) were retrospectively evaluated in consensus by 2 radiologists. All studies were evaluated for the presence of multiple imaging features. Two radiographic scores were then calculated: a total radiographic score and a radiographic active inflammation score. In patients who underwent MD-CTE, pouch endoscopy, and biopsy within 30 days (n = 13), both scores were correlated with findings on pouch endoscopy and histopathology. RESULTS: Of the 35 patients, 33 (94%) had at least one MD-CTE finding of active or chronic pouch inflammation and 27 patients (77%) had at least one MD-CTE finding of active pouch inflammation. Of the 13 patients who underwent endoscopy and biopsy, the total radiographic score demonstrated a strong positive correlation with endoscopic score (r = 0.81; P = 0.001) and a moderate positive correlation with histopathologic score (r = 0.56; P = 0.047). The radiographic active inflammation score demonstrated a strong positive correlation with endoscopic score (r = 0.83; P = 0.0004), but only a weak nonsignificant positive correlation with histopathologic score (r = 0.492, P = 0.087). CONCLUSIONS: In patients who had IPAA surgery, findings on MD-CTE correlate positively with findings on pouch endoscopy and histopathology and are sensitive measures for pouch inflammation with high positive predictive value. Thus, MD-CTE can be a useful noninvasive test in the early evaluation of symptomatic patients.

MDCT assessment of ulcerative colitis: radiologic analysis with clinical, endoscopic, and pathologic correlation.

PURPOSE: Evaluate the utility of multidetector-row computed tomography (MDCT) in assessing the severity of ulcerative colitis (UC) in comparison with clinical assessment, colonoscopy, and histopathology. MATERIALS AND METHODS: Patients with UC evaluated with at least one abdominal contrast-enhanced CT study (CECT) within 7 days of colonoscopy with biopsy were included. CECT of 23 patients (12 male; mean age 40 years; age range, 20-72 years) were retrospectively evaluated in consensus by two radiologists. A total of 138 lower GI tract segments were evaluated by CECT and graded for the presence of bowel wall thickening, mucosal hyperenhancement, mural stratification, mesenteric hyperemia, pericolonic stranding, and lymph nodes. A cumulative CT severity score was calculated and correlated with clinical, colonoscopic, and histopathologic severity grades. RESULTS: The cumulative CT score and individual CECT scores for bowel wall thickening, mucosal hyperenhancement, and mural stratification showed positive correlation with clinical severity (P < 0.05). All individual CECT features as well as the cumulative CT score demonstrated statistically significant correlation with colonoscopic severity (P < 0.0001). Only wall thickening on CECT demonstrated significant correlation with histopathologic severity (P = 0.01). CONCLUSION: Disease severity assessment by MDCT demonstrates positive correlation with severity established by clinical assessment and colonoscopy. Only increasing wall thickness, as graded on MDCT, correlates with histopathologic disease severity.

Diffusion-weighted MRI of focal cystic pancreatic lesions at 3.0-Tesla: preliminary results.

PURPOSE: To determine the utility of 3.0-Tesla diffusion-weighted (DW) magnetic resonance imaging (MRI) for focal cystic pancreatic lesion (FCPL) characterization. MATERIALS AND METHODS: 55 FCPL (34 IPMN, 5 serous cystadenoma, and 16 inflammatory) were evaluated. Two radiologists reviewed in consensus DW-MRI images. Reference standard was obtained from patient history, cytological and histopathology data, FCPL fluid analysis, and follow-up imaging results. Signal intensity (SI) and apparent diffusion coefficient values (ADC) of FCPL and normal pancreas were measured. FCPL-to-pancreas SI and ADC ratios were also calculated. RESULTS: Qualitatively, 11 of 21 non-mucinous vs. 4 of 34 mucinous lesions appeared hyperintense at b value of 1,000 s/mm(2) (P = 0.02). Three FCPL demonstrated restricted diffusion: all inflammatory. Significant differences in mean ADC between neoplastic vs. non-neoplastic (P = 0.009), and mucinous vs. non-mucinous (P = 0.013) lesions were demonstrated. FCPL-to-pancreas ADC and SI ratios demonstrated significant differences between neoplastic vs. non-neoplastic lesions [ADC, (P = 0.019); SI for b values 750 (P = 0.010) and 1,000 s/mm(2) (P = 0.017)] and mucinous vs. non-mucinous lesions [ADC (P = 0.018); SI for b values 750 (P = 0.013) and 1,000 s/mm(2) (P = 0.015)]. CONCLUSION: Although mean ADC values and FCPL-to-pancreas SI and ADC ratios may be helpful in differentiating FCPL, characterization of individual FCPL by means of 3.0-Tesla DW-MRI appears limited.

Dynamic MR defecography: assessment of the usefulness of the defecation phase.

OBJECTIVE: The purpose of this study was to assess the usefulness of the defecation phase during dynamic MR defecography. MATERIALS AND METHODS: The images from 85 MR defecographic examinations (83 patients; age range, 20-88 years; mean, 52.7) were retrospectively reviewed in consensus by two observers. Images from each of four phases (rest, maximal sphincter contraction and squeezing, maximal straining, and defecation) were evaluated and scored independently with a modified previously published grading system. Features evaluated included the presence and degree of bladder, vaginal, and rectal descent and the presence and size of rectocele, enterocele, and intussusception. Statistical analysis was performed with a variety of tests. RESULTS: Compared with images obtained in the other phases, defecation phase images helped in identification of additional cases of abnormal bladder descent in 43 examinations (50.6%), abnormal vaginal descent in 52 examinations (61.2%), and abnormal rectal descent in 11 examinations (12.9%). Similarly, only defecation phase images depicted previously undetected rectoceles 2 cm or larger in 31 examinations (36.5%), enteroceles in 34 examinations (40%), and intussusceptions in 22 examinations (25.9%). The number of additional cases of abnormalities identified on defecation phase images was significantly greater than the number identified on images obtained in the other phases (p < 0.005). The average total scores for the rest, squeeze, strain, and defecation phases were 1.4, 0.7, 2.3, and 6.6. The average total defecation phase score was significantly greater than the average total score in any of the other phases (p < 0.001). CONCLUSION: During dynamic MR defecography, defecation phase imaging yields important additional information on the presence and degree of pelvic floor abnormalities and is therefore an essential component of MR defecographic examinations.

Feasibility of low-tube-voltage excretory phase images during CT urography: assessment using a dual-energy CT scanner.

OBJECTIVE: The purpose of this study is to assess the feasibility of low-tube-voltage images during excretory phase CT urography. MATERIALS AND METHODS: In this retrospective study, we examined 70 consecutive CT urograms (35 men and 35 women; mean age, 58.5 years) performed on a dual-energy CT scanner and compared excretory phase images obtained at 80 kVp and 340 mAs with blended images (0.3 × 140 kVp and 80 mAs; and 0.7 × 80 kVp and 340 mAs). Quantitative measurements of urinary system opacification (Hounsfield units), image noise (Hounsfield units), and effective dose (millisieverts) were compared using Student paired t test. Image noise was correlated with patient thickness. Two independent blinded readers qualitatively assessed opacification, image quality (both compared using Wilcoxon test), overall acceptability (compared using McNemar test), and detectability of urinary and extraurinary findings. RESULTS: The 80-kVp images yielded significantly higher opacification of renal pelvis (p < 0.0001), ureter (p < 0.0001), bladder (p < 0.0001), and aorta (p < 0.0001); higher image noise (p < 0.0001); and lower radiation dose (5.2 vs 11.9 mSv). Image noise increased along with increasing patient thickness (r = 0.86 for 80-kVp images). Qualitative opacification scores were better only in the bladder on 80-kVp images (p = 0.002). Although 80-kVp image quality was lower (p < 0.0001), the overall acceptability was similar. Of 42 urinary findings, 40 were detected on 80-kVp images (< 2-mm calyceal calculus and tiny foci of collecting system gas were missed in one patient each, both large patients). Of 137 extraurinary findings, 130 were detected on 80-kVp images (no findings of high clinical significance were missed). CONCLUSION: Low tube voltage (80 kVp) during excretory phase CT urography is feasible, with improved urinary system opacification, acceptable image quality, and lower radiation dose.

The radiological management of the thrombosed arteriovenous dialysis fistula.

Patent vascular access is a prerequisite for adequate haemodialysis, and is a major determinant of quality of life and long-term survival of patients with end-stage renal disease. Autogenous haemodialysis fistulas (AVFs) have demonstrated superior clinical outcome when compared to synthetic grafts, but both types of access remain susceptible to venous stenoses, and consequent thrombotic occlusion. Recent publications have reported primary patency rates of up to 100% following percutaneous de-clotting of AVFs incorporating techniques such as pharmacological thrombolysis, mechanical thrombectomy, and thrombo-aspiration. Endovascular management also provides information regarding the underlying cause of access thrombosis, with option to treat. Consequently, there has been a paradigm shift in the management of fistula thrombosis, with interventional radiology assuming a lead role in initial salvage procedures. This article will attempt to provide the reader with an insight into the multiple radiological techniques that can be employed to salvage a thrombosed AVF based on current published literature.

Usefulness of percutaneous biopsy in diagnosing benign renal masses that mimic malignancy.

Percutaneous biopsy has long been used to diagnose malignancies of the kidney. It is an established technique with multiple indications. Percutaneous biopsy now can be used to diagnose benign conditions that may mimic a malignancy and lead to unnecessary treatments. Advances in cytological techniques such as immunocytochemistry and cytogenetics have allowed for an increased diagnostic yield. In this review, various benign entities that may present as a renal mass are discussed and the vital role of percutaneous biopsy detailed.

SARS-CoV-2 Detection in Gingival Crevicular Fluid.

Understanding the pathophysiology of the coronavirus disease 2019 (COVID-19) infection remains a significant challenge of our times. The gingival crevicular fluid being representative of systemic status and having a proven track record of detecting viruses and biomarkers forms a logical basis for evaluating the presence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The study aimed to assess gingival crevicular fluid (GCF) for evidence of SARS-CoV-2 in 33 patients who were deemed to be COVID-19 positive upon nasopharyngeal sampling. An attempt was also made to comparatively evaluate it with saliva in terms of its sensitivity, as a diagnostic fluid for SARS-CoV-2. GCF and saliva samples were collected from 33 COVID-19-confirmed patients. Total RNA was extracted using NucliSENS easyMAG (bioMérieux) and eluted in the elution buffer. Envelope gene (E gene) of SARS-CoV-2 and human RNase P gene as internal control were detected in GCF samples by using the TRUPCR SARS-CoV-2 RT qPCR kit V-2.0 (I) in an Applied Biosystems 7500 real-time machine. A significant majority of both asymptomatic and mildly symptomatic patients exhibited the presence of the novel coronavirus in their GCF samples. Considering the presence of SARS-CoV-2 RNA in the nasopharyngeal swab sampling as gold standard, the sensitivity of GCF and saliva, respectively, was 63.64% (confidence interval [CI], 45.1% to 79.60%) and 64.52% (CI, 45.37% to 80.77%). GCF was found to be comparable to saliva in terms of its sensitivity to detect SARS-CoV-2. Saliva samples tested positive in 3 of the 12 patients whose GCF tested negative, and likewise GCF tested positive for 2 of the 11 patients whose saliva tested negative on real-time reverse transcription polymerase chain reaction. The results establish GCF as a possible mode of transmission of SARS-CoV-2, which is the first such report in the literature, and also provide the first quantifiable evidence pointing toward a link between the COVID-19 infection and oral health.

Mixed epithelial and stromal tumour of the kidney: imaging features.

OBJECTIVE: To describe the features on ultrasonography (US), computed tomography (CT) and magnetic resonance imaging (MRI) of mixed epithelial and stromal tumours of the kidney. PATIENTS AND METHODS: Five women with renal mixed epithelial and stromal tumours (mean age 55.6 years, range 49-59) who had preoperative imaging were retrospectively analysed. Three ultrasonograms, five contrast-enhanced CT scans, and one contrast-enhanced MRI examination were available for review. Specific imaging features analysed included lesion size, location, enhancement and cystic composition. The presence of calcification, septation, nodularity and a capsule were also evaluated. RESULTS: All mixed epithelial and stromal tumours appeared as well-marginated, multi-septate cystic masses with a nodular component. All lesions were classified as Bosniak category III (three) or IV (two). The presence of calcification (four) or a capsule (three) was variable. Two tumours invaginated into the renal pelvis with no invasion. CONCLUSION: Mixed epithelial and stromal tumours of the kidney have a diverse radiographic appearance, indistinguishable from multilocular cystic nephroma and cystic renal cell carcinoma.

Autoimmune hepatitis: CT and MR imaging features with histopathological correlation.

BACKGROUND: To describe the CT and MRI features of autoimmune hepatitis (AIH) and correlate them with histological grade and stage. Observed changes associated with treatment are also described. METHODS: A retrospective analysis of the initial CT scans (n = 22) and MRI exams (n = 12) of 27 patients with pathologically-proven AIH was conducted. Multiple objective and subjective imaging features were evaluated. Correlation of imaging features with histological inflammatory grade and fibrotic stage was performed using the Fisher exact test and Spearman's rank correlation coefficient. In eight patients serial CT and MR imaging during treatment was used to describe the changes associated with treatment. RESULTS: The presence of ascites, expanded gallbladder fossa, spleen size, and enlarged preportal space had significant positive correlations with fibrotic stage. No significant positive correlations existed between imaging features and portal or lobular inflammatory grade. Seven patients (25.9%) were normal. The most common abnormal finding was surface nodularity: CT (n = 11 [50%]) and MRI (n = 8 [66.7%]). There was a wide variability in imaging appearances of patients who had serial scans on treatment. CONCLUSIONS: There is a wide spectrum of CT and MR imaging features in patients with AIH. Several MRI features demonstrate a significant positive correlation with fibrotic stage.