Usefulness of CT texture analysis in differentiating benign and malignant renal tumours

AIM

To elucidate visually imperceptible differences between benign and malignant renal tumours using computed tomography texture analysis (CTTA) using filtration histogram based parameters.

MATERIALS AND METHODS

A retrospective study was performed by texture analysis of pretreatment contrast-enhanced CT examinations in 354 histopathologically confirmed renal cell carcinomas (RCCs) and 147 benign renal tumours. A region-of-interest was drawn encompassing the largest cross-section of the tumour on venous phase axial CT. CTTA features of entropy, kurtosis, mean positive pixel density, and skewness at different spatial filters were calculated and compared in an attempt to differentiate benign lesions from malignancy.

RESULTS

Entropy with fine spatial filter was significantly higher in RCC than benign renal tumours (p=0.022). Entropy with fine and medium filters was higher in RCC than lipid-poor angiomyolipoma (p=0.050 and 0.052, respectively). Entropy >5.62 had high specificity of 85.7%, but low sensitivity of 31.3%, respectively, for predicting RCC.

CONCLUSIONS

Differences in entropy were helpful in differentiating RCC from lipid-poor angiomyolipoma, and chromophobe RCC from oncocytoma. This technique may be useful to differentiate lesions that appear equivocal on visual assessment or alter management in poor surgical candidates.

Imaging features of haematological malignancies of kidneys

Haematological malignancies are relatively uncommon neoplasms of kidneys. Nevertheless, the incidence of these neoplasms is increasing, partly due to more widespread use of computed tomography and magnetic resonance imaging. This article discusses the clinical and imaging features of renal lymphoma, leukaemia, extra-osseous multiple myeloma, and post-transplant lymphoproliferative disorder. Although there is overlap of imaging features with other more common malignancies, such as transitional and renal cell cancers, the combination of imaging findings and the appropriate clinical picture should allow the radiologist to raise a provisional diagnosis of a haematological neoplasm. This has management implications including the preference for image-guided core biopsies and a shift towards medical rather than surgical therapy.

The usefulness of diffusion-weighted imaging in the characterization of liver lesions in patients with cirrhosis

AIM

To evaluate if diffusion-weighted imaging (DWI) is useful in characterizing liver lesions in patients with cirrhosis.

MATERIALS AND METHODS

A retrospective review revealed 37 patients with cirrhosis who had 41 histologically proven hepatocellular carcinoma (HCC) lesions. Another 20 patents with cirrhosis had 29 solid nodules that remained stable for at least 12 months and were deemed to be benign hepatic nodules (BHN). Of the HCC lesions, 14 were well-differentiated (WD HCC), 20 were moderately differentiated, and seven were poorly differentiated histology. For all lesions, two reviewers analysed signal characteristics and made apparent diffusion coefficient value (ADC) measurements.

RESULTS

Visual analysis of DWI was useful in that no HCC was hypointense and no BHN was hyperintense to liver. Visual analysis of DWI was not useful in separating WD HCC from higher grades. There was substantial overlap in ADC values of the HCC and BHN. Among HCC lesions, ADC values of more than 0.99 × 10(-3) mm(2)/s had sensitivity and specificity of 85% and 86% for reviewer 1, and 63% and 64% for reviewer 2 in diagnosing WD HCC.

CONCLUSIONS

ADC measurements of BHN were higher than that of HCC, and the ADC values of WD HCC were higher than that of more aggressive grades of HCC. However, quantitative measurements may not help in determining the histological grade of individual cases of HCC.

Accuracy of magnetic resonance cholangiopancreatography in the diagnosis of pancreas divisum

BACKGROUND

Patients with pancreas divisum may develop pancreatitis. Endoscopic retrograde cholangiopancreatography (ERCP) is the gold standard for diagnosing pancreas divisum. Magnetic resonance cholangiopancreatography (MRCP) is a non-invasive test reported to be highly accurate in diagnosing pancreas divisum.

AIM

To evaluate the diagnostic accuracy of MRCP in detecting pancreas divisum at our institution.

METHODS

We reviewed patients who underwent both ERCP and MRCP. Patients who had diagnostic endoscopic pancreatograms (ERP) after MRCP comprise the study population. Secretin was given in 113/146 patients (S-MRCP). The remaining 33/146 patients had MRCP without secretin. In 7/33 patients who underwent MRCP without secretin (21.2%), the studies were non-diagnostic and, therefore, this group was not further analyzed and the study focused on the S-MRCP group only.

RESULTS

ERP identified pancreas divisum in 19/113 (16.8%) patients. S-MRCP identified 14/19 pancreas divisum and was false-positive in three cases (sensitivity 73.3%, specificity 96.8%, positive predictive value 82.4%, negative predictive value 94.8%). Of the eight patients with inaccurate S-MRCP, 5 (63%) had changes of chronic pancreatitis by ERP. This differs from the frequency of chronic pancreatitis by ERP in 24/105 (23%) patients with accurate MRCP findings. The ERCP findings of chronic pancreatitis were more frequent among incorrect S-MRCP interpretations than among correct interpretations (odds ratio [OR] 5.5 [95% confidence interval (CI) 1.3-25.3]). MRCP without secretin is non-diagnostic for pancreas divisum in a significant proportion of patients. S-MRCP had a satisfactory specificity for detecting pancreas divisum. However, the sensitivity of S-MRCP for the diagnosis of pancreas divisum was modest at 73.3%. This is low compared to previous smaller studies, which reported a sensitivity of MRCP of up to 100%.

Does the hepatocellular phase of gadobenate dimeglumine help to differentiate hepatocellular carcinoma in cirrhotic patients according to histological grade?

AIM

To assess the role of the hepatocellular phase on magnetic resonance imaging (MRI) following gadobenate in characterizing the grade of hepatocellular carcinoma (HCC) in cirrhotic patients.

MATERIALS AND METHODS

A retrospective review of the MRI database from October 2004 to February 2009, performed for this Institutional Review Board-approved and Health Insurance Portability and Accountability Act (HIPAA)-complaint study, revealed 237 cirrhotic patients with focal liver lesions. Patients who had both a hepatocellular phase after gadobenate and pathological confirmation of HCC were included. Forty-six patients with 73 HCC were analysed independently by two reviewers for signal characteristics. Absolute contrast-to-noise ratio (CNR) and enhancement ratio (ER) were calculated. Univariate analysis, stepwise logistic regression analysis, and receiver operating characteristic curves (ROC) were performed.

RESULTS

The mean age was 61.3 years (range 45 to 78 years). There were 11 females and 35 males, who had 22 well-differentiated (WD HCC), 35 moderately-differentiated (MD HCC), and 16 poorly-differentiated (PD HCC) hepatocellular carcinomas. On visual analysis of the hepatocellular phase, a hyperintense or isointense lesion had a sensitivity and specificity of 45% and 76%, respectively, for WD HCC. On quantitative analysis, the only significant predictor of the grade of HCC was the ER on the hepatocellular phase (p=0.019 and 0.001 for the two reviewers in logistic regression model). On ROC analysis, an ER of >13% was 47% sensitive and 89% specific in predicting WD HCC histology.

CONCLUSION

Although the hepatocellular phase of gadobenate may help to differentiate some cases of WD HCC from the more aggressive grades, there is overlap between the different grades on qualitative and quantitative analysis.

Diffusion-weighted imaging in characterization of cystic pancreatic lesions

AIM

To evaluate whether apparent diffusion coefficient (ADC) measurements from diffusion-weighted imaging (DWI) can characterize or predict the malignant potential of cystic pancreatic lesions.

MATERIALS AND METHODS

Retrospective review of the magnetic resonance imaging (MRI) database over a 2-year period revealed 136 patients with cystic pancreatic lesions. Patients with DWI studies and histological confirmation of cystic mass were included. In patients with known pancreatitis, lesions with amylase content of >1000 IU/l that resolved on subsequent scans were included as pseudocysts. ADC of cystic lesions was measured by two independent reviewers. These values were then compared to categorize these lesions as benign or malignant using conventional MRI sequences.

RESULTS

Seventy lesions were analysed: adenocarcinoma (n=4), intraductal papillary mucinous neoplasm (IPMN; n=28), mucinous cystic neoplasm (MCN; n=9), serous cystadenoma (n=16), and pseudocysts (n=13). There was no difference between ADC values of malignant and non-malignant lesions (p=0.06), between mucinous and serous tumours (p=0.12), or between IPMN and MCN (p=0.42). ADC values for low-grade IPMN were significantly higher than those for high-grade or invasive IPMN (p=0.03).

CONCLUSION

ADC values may be helpful in deciding the malignant potential of IPMN. However, they are not useful in differentiating malignant from benign lesions or for characterizing cystic pancreatic lesions.

Capsule endoscopy and imaging tests in the elective investigation of small bowel disease

Capsule endoscopy enables high-resolution depiction of small bowel mucosa and has been shown, by several studies, to have a high diagnostic yield in a variety of small bowel diseases. In this review, we critically assess the contributions of capsule endoscopy and imaging tests in common small bowel disorders. Radiological tests that only assess the small bowel mucosa will be less useful in the era of capsule endoscopy.

Positron-emission tomography and computed tomography of cystic pancreatic masses

AIM

To investigate the sensitivity and specificity of computed tomography (CT), positron-emission tomography (PET), and both methods in combination, for determining whether cystic pancreatic tumours are malignant.

MATERIALS AND METHODS

We retrospectively identified all patients with cystic pancreatic tumours who underwent separate PET and contrast-enhanced CT examinations within a 1-month interval. Tumours were classified as benign or malignant on CT (two radiologists, independently), PET [a reported standardized uptake value (SUV) of 2.5 was taken as the cut-off between benign and malignant], and with PET and CT images together (two radiologists, in consensus). Readers were blinded to pathological and other radiological findings. Mean patient age and lesion size were compared between benign and malignant groups using Student's t-test. For CT findings, odds ratios (OR) and confidence intervals (CI) were calculated using multivariate logistic regression models. For CT, PET, and the combined images, sensitivities and specificities were calculated, and compared between groups using Fisher's exact test.

RESULTS

Thirty patients were identified. The best CT predictor of malignancy was size; mean diameter was 2.3 cm (benign) and 4.1 cm (malignant) (p<0.01); OR was 2.80 (95% CI, 1.26-6.20). Sensitivities of CT, PET and combined PET/CT images were 67-71, 57, and 86%, respectively. PET/CT was more sensitive than PET (p<0.01) or CT (p<0.01) alone. Specificities of CT, PET, and combined PET/CT images were 87-90, 65, and 91%, respectively. PET/CT was more specific than PET (p<0.01) but not CT (p>0.05).

CONCLUSION

The sensitivity and specificity of combined PET and CT images is comparable with or superior to either CT or PET alone in determining malignancy in cystic pancreatic lesions.

Benefits of routine use of coronal and sagittal reformations in multi-slice CT examination of the abdomen and pelvis

AIM

To evaluate the usefulness of coronal and sagittal reformations from isotropic abdomino-pelvic computed tomography (CT) examinations.

METHODS

Fifty consecutive abdomino-pelvic CT examinations were reconstructed into two sets of axial source images: 0.9 mm section width with 0.45 mm reconstruction interval (isotropic) and 4 mm section width with 3 mm reconstruction interval. The isotropic dataset was reformatted into coronal and sagittal stacks with 4 mm section widths. Three readers independently reviewed the three image sets with 4 mm section widths. The coronal and sagittal reformations were compared with the axial images, in the same sitting, for depiction of lesions in various abdominal organs.

RESULTS

There was better visualization of lesions in the liver, kidneys, mesentery, lumbar spine, major abdominal vessels, urinary bladder, diaphragm and hips on the coronal reformations compared with source axial images (p<0.05). Sagittal reformations scored better than axial source images for showing lesions in the liver, thoracic spine, abdominal vessels, uterus, urinary bladder, diaphragm and hips (p<0.05). The coronal and sagittal series showed significant additional information in 23 and 17% of patients, respectively.

CONCLUSION

Radiologists should consider the routine review of at least one additional plane to the axial series in the interpretation of abdomino-pelvic CT studies.

Chemotherapy-induced sclerosing cholangitis

AIM

To review the computed tomography (CT), magnetic resonance imaging (MRI) and cholangiographic findings of chemotherapy-induced sclerosing cholangitis (CISC).

METHODS

Between January 1995 and December 2004, 11 patients in the endoscopic retrograde cholangiography database were identified with CISC. Twelve CT, four MRI, 69 endoscopic and nine antegrade cholangiographic studies in these patients were reviewed. Serial change in appearance and response to endoscopic treatment were recorded.

RESULTS

CISC showed segmental irregular biliary dilatation with strictures of proximal extrahepatic bile ducts. The distal 5cm of common bile duct was not affected in any patient. CT and MRI findings included altered vascular perfusion of one or more liver segments, liver metastases or peritoneal carcinomatosis. Biliary strictures needed repeated stenting in 10 patients (mean: every 4.7 months). Cirrhosis (n=1) or confluent fibrosis (n=0) were uncommon findings.

CONCLUSION

CISC shares similar cholangiographic appearances to primary sclerosing cholangitis (PSC). Unlike PSC, biliary disease primarily involved ducts at the hepatic porta rather than intrahepatic ducts. Multiphasic contrast-enhanced CT or MRI may show evidence of perfusion abnormalities, cavitary liver lesions, or metastatic disease.

Isotropic chest CT examination: diagnostic quality of reformats

AIM

To evaluate the image quality of axial and coronal reformats obtained from isotropic resolution 40-channel chest computed tomography (CT) examinations.

METHODS AND MATERIALS

Thirty consecutive patients with intravenous contrast-enhanced chest CT examinations using a 40-channel CT machine were enrolled for the study. The raw data were reconstructed into two sets of source axial images: 0.9 mm section width and 0.45 mm reconstruction interval (isotropic resolution) and 4mm section width with 3 mm reconstruction interval (anisotropic resolution; group A). The isotropic data set was reformatted into axial and coronal stacks (groups B and C, respectively) with 4 mm section width and 3 mm interval. Three independent readers evaluated stacks A to C using a three-point scale for resolution of right lower lobe segmental bronchi, edge sharpness of major and minor fissures, respiratory motion artefact, reconstruction artefact, noise and overall image quality.

RESULTS

The sharpness of fissures scored significantly higher with the coronal reformats (group C) compared with the axial image sets (groups A and B) (p<0.01). Noise in group A scored significantly lower than groups B or C (p<0.01). For other parameters there was no statistical difference between the groups. There was substantial or excellent agreement between the reviewers.

CONCLUSION

Isotropic imaging of the chest allows creation of reformats with similar image quality as similar thickness axial source images. These reformats are probably of sufficient quality to form the basis of clinical interpretation.

Routine isotropic computed tomography scanning of the abdomen and pelvis

For 30 years, abdominal CT has been imaged and reviewed in the axial plane. It is now possible to carry out isotropic imaging of the whole abdomen and pelvis using a 40-channel scanner. This allows creation of coronal and sagittal reformats with the same image quality as the axial images. In this study, we present our experience of reviewing routinely coronal and, occasionally, sagittal reformats. We discuss situations where these nonaxial reformats are most beneficial.

Enteric drainage pancreatic transplantation

Enteric drainage is currently the preferred method of pancreatic transplantation. This technique results in long-term good control of diabetes. In this report we discuss the postoperative radiologic anatomy and complications.

Hepatic peliosis (bacillary angiomatosis) in AIDS: CT findings

Patients with acquired immunodeficiency syndrome are at risk of developing opportunistic infections and aggressive tumors. Computed tomographic examination is the usual method of evaluating the abdomen and pelvis in these patients. Although this technique is reasonably sensitive in detecting pathology, findings are often nonspecific. A case of hepatic peliosis (bacillary angiomatosis) in a patient with acquired immunodeficiency syndrome is presented.

Mesenteric desmoid mimicking recurrent testicular cancer

The detection of a solid abdominal mass after surgery for testicular cancer most frequently represents metastatic disease. We report an unusual case of mesenteric desmoid tumor presenting after retroperitoneal lymph node dissection for metastatic testicular cancer. To our knowledge, there are no similar reports in the radiology literature.

Radiology of small bowel obstruction: contemporary approach and controversies

The radiologic workup of patients with known or suspected small bowel obstruction and the timing of surgical intervention in this complex situation have undergone considerable changes over the past two decades. The diagnosis and treatment of small bowel obstruction, a common clinical condition often associated with signs and symptoms similar to those seen in other acute abdominal disorders, continue to evolve. This article examines the changes related to the use of imaging in the diagnosis and management of patients with this potentially dangerous problem and revisits pertinent controversies.

Proximal small bowel intussusceptions in adults: CT appearance and clinical significance

BACKGROUND

According to the clinical literature, intestinal intussusception in adults is rare, is associated with a pathologic lead point, and is usually treated surgically. Nonobstructing small bowel intussusception has been reported as a transient finding on computed tomographic (CT) studies.

METHODS

We evaluated the radiographic and clinical findings in 24 patients who were found to have 26 proximal small bowel intussusceptions on abdominal CT scans performed for a variety of indications.

RESULTS

Twenty patients with intussusceptions had no evidence of small bowel obstruction. The transient and clinically insignificant nature of 22 intussusceptions in these 20 patients was proven radiologically (n = 14), surgically (n = 1), or by clinical follow-up (n = 7). These patients demonstrated a bowel-within-bowel pattern on multiple contiguous images and absence of strangulation or intestinal dilatation. No lead points were demonstrated in these patients. Three other patients had symptoms of low-grade small bowel obstruction and were treated conservatively. Extensive follow-up investigations showed no recurrence of intussusception or a lead point. One patient had high-grade obstructive intussusception with intestinal ischemia and required surgical resection of necrotic bowel.

CONCLUSIONS

Proximal small bowel intussusceptions are likely to be transient and nonobstructive and unlikely to have a significant lead point.

Temporal bone: comparison of isotropic helical CT and conventional direct axial and coronal CT

OBJECTIVE

The objective of the study was to compare helical CT (with reformation of coronal images from the axial data set) with conventional direct axial and coronal CT of the temporal bones.

SUBJECTS AND METHODS

Nineteen patients underwent both conventional 1-mm direct axial and coronal CT and helical 0.5-mm axial CT. The helical data set was reconstructed at 0.2-mm increments, and axial and coronal images were reconstructed in a plane similar to that of the conventional study, with a slice width of 0.5 mm and 0.5-mm increments. Forty small structures were evaluated independently by three observers, who were unaware of the method of imaging. Observers graded the 40 structures using a modified Likert scale. The graded differences between the two techniques were evaluated using a paired t test. Correlation between observers' gradings was evaluated using analysis of variance.

RESULTS

The helical CT technique scored significantly higher than the conventional technique for many individual structures and groups of structures (scutum [p = .041], stapes footplate [p = .006], stapes crura [p = .004], oval window [p = .026], crista falciformis [p = .006], whole temporal bone [P = .012], middle ear [p = .033], inner ear [p = .021], ossicles [p = .044], and stapes [p = .010]). The correlation coefficient among observers was .91 for the whole temporal bone.

CONCLUSION

Helical CT using 0.5-mm technique and reconstruction produces diagnostic images comparable with or superior to conventional 1-mm technique because helical CT can obtain thinner slices.

Measurement of liver volume using spiral CT and the curved line and cubic spline algorithms: reproducibility and interobserver variation

BACKGROUND

The purpose of this article was to evaluate the accuracy and reproducibility of spiral computed tomography (CT) and the curved line and cubic spline algorithms in measuring liver volume.

METHODS

Spiral CT was performed in phantoms, cadaveric liver specimens, and 35 live human subjects (19 healthy volunteers and 16 patients). Images were transferred to a workstation, and volumes were measured by two observers. One observer repeated the measurements at a separate sitting.

RESULTS

The correlation between the CT measurement and the gold standard measurement of the cadaveric livers was very strong (r = 0.94). For the live human subjects, the intraobserver and interobserver correlations were extremely high (r = 0.999 and 0.997, respectively). The mean difference in liver volume measurements between the separate observations was 1%.

CONCLUSION

The accuracy and reproducibility of this method of assessing liver volume are very high.

Gender differences in alcohol metabolism: relationship to liver volume and effect of adjusting for body mass

BACKGROUND & AIMS

Alcoholic liver disease purportedly develops more readily in women than in men. Some studies have demonstrated faster rates of alcohol elimination in women. This study examined whether gender differences in alcohol metabolism are related to differences in liver volume and/or differences in lean body mass.

METHODS

Ten men and 10 women had alcohol elimination rates determined by clamping of the breath alcohol concentration at 50 mg/dL by means of a constant rate of intravenous infusion of 6% ethanol. Liver volume was determined by computed tomography.

RESULTS

Mean alcohol elimination rate and mean computed liver volume were not significantly different in men and women. Lean body mass was 42% greater in men than in women. Consequently, the calculated alcohol elimination rate and liver volume per kilogram of lean body mass were 33% and 38% higher in women than in men, respectively. When the alcohol elimination rate was calculated per unit liver volume, no gender-related difference was found.

CONCLUSIONS

Women have greater clearance of ethanol per unit lean body mass, confirming previous oral alcohol administration studies. Women have approximately the same liver volume as men, explaining the equivalent alcohol elimination rates seen when men and women are compared on the basis of liver size.

Postinfective diarrhoea and bile acid malabsorption

Postinfective irritable bowel syndrome with diarrhoea and idiopathic bile acid malabsorption remains an enigma. We examined the records of 84 patients whose 75SeHCAT scans were indicative of bile acid malabsorption (< 15% one week retention). Identifiable causes of bile acid malabsorption were: previous ileal surgery (7), Crohn's disease (22), radiation enteritis (13), vagotomy, gastrectomy or cholecystectomy (10) and miscellaneous (3). Sixteen of 29 patients with apparently idiopathic bile acid malabsorption gave a clear history of acute gastroenteritis before the onset of chronic diarrhoea lasting from 0.25-18 years until their positive 75SeHCAT scan. Only four cases of campylobacter, and one each of shigella and salmonella were documented. Extensive investigation failed to detect other possible pathologies. In response to bile acid sequestrants, mean stool frequency fell from 7.2 per day to 2.1 per day (p < 0.001). We have observed that postinfective chronic diarrhoea is associated with chronic bile acid malabsorption, which can be successfully treated with bile acid sequestrants such as cholestyramine.

Magnetic resonance imaging of septic sacroiliitis

Five cases of septic sacroiliitis diagnosed by magnetic resonance imaging (MRI) are presented. Imaging was performed between 2 and 14 days after onset of symptoms and consisted of varying combinations of coronal short tau inversion recovery (STIR), axial T2-weighted spin echo (SE), and coronal and axial pre- and postcontrast T1-weighted SE scans. Abnormalities included demonstration of sacroiliac joint effusions, bone oedema and adjacent inflammation as high signal on STIR and T2-weighted SE scans, and identification of abscesses in two cases as rim-enhancing lesions anterior to the joint on gadolinium-enhanced T1-weighted SE scans. The role of MRI and other forms of imaging in septic sacroiliitis is discussed.

Sub-clinical systemic lupus erythematosus presenting with acute myocarditis

A 46 year old woman presented with fever and normochromic anaemia followed rapidly by severe myocardial failure, unresponsive to maximum inotropic support and broad spectrum antibiotics. There were no classical clinical stigmata of systemic lupus erythematosus (SLE) but a possible immunological cause was looked for, and on the basis of her immuno-serology a diagnosis of SLE-like disease was made. She responded rapidly to high dose steroids. The importance of considering the possibility of SLE or 'lupus overlap' in an acutely ill 'undiagnosed' patient is emphasized. The relevance of instigating appropriate immuno-serological tests in the course of such an illness is discussed.