| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||
ABSTRACT |
Thursday, May 4, 1:30 PM-3:30 PM
Abstracts 273-284
Moderators: James Ellis, MD and Deborah A. Baumgarten, MD
1:30 PM
273. Validation of Deconvolution-based MR Imaging of Renal Perfusion using Standard Gd-chelates in Healthy Volunteers
Michaely H.J.1*; Dujardin M.2; Sourbron S.2; Luypaert R.2; Reiser M.F.1; Schoenberg S.O.1; 1. Institute of Clinical Radiology, University of Munich - Grosshadern Campus, Munich, Germany; 2. Radiology/BEFY, Vrije Universiteit Brussel, Brussel, Belgium.
Address correspondence to H.J. Michaely (henrik.michaely{at}med.uni-muenchen.de)
Objective: Previously we applied a deconvolution approach for the quantification of perfusion to renal DCE-MRI data, and observed significant underestimations with respect to gold-standards. Here we examine whether the use of an optimized MRI sequence reduces such systematic errors.
Materials and Methods: Perfusion measurements were performed in 9 male volunteers (25-35 years) without renal disease. On 1.5T (Siemens Avanto), four slices were acquired per second using SR-TurboFLASH (TR 254 ms, TE 1.04 ms, FA 12°, TI 13 ms, slice thickness 8 mm) after injection of 7 mlGd-BOPTA (Multihance) at 4ml/s. The relative signal enhancement was measured from tissue-time-curves (TTC). An impulse response function (IRF) was obtained by deconvolving the TTCs with an AIF selected manually in the aorta on the axial slice. Parametric maps of relative renal blood flow (rRBF) and relative renal volume of distribution (rRVD) were calculated as maximum and time-integral of the IRF, and relative mean transit time (rMTT) as the ratio rRVD/rRBF. Average values for single kidney cortex were compared to literature values for normal renal perfusion.
Results: Parametric images and values could be obtained in all patients. Cortical areas could easily be delineated from medulla. Average rRBF, rRVD and rMTT were 308.1 ml/min/100 ml ± 50.0, 36.0 ml/100 ml ± 7.9 and 7.3s ± 1.2, respectively. rRBF values are in agreement with existing gold standards.
Conclusion: The results are consistent with the hypothesis that errors in previous results were due to inflow effects in the data. The use of the improved sequence removes this systematic error without compromising image quality.
274. Comparison of 2D-SR-TurboFLASH with and without Parallel Imaging with 2D-SR-TrueFISP for Semiquantitative Assessment of First Pass Renal Perfusion at 1.5T
Michaely H.J.1,3*; Nael K.3; Kramer H.1; Lodemann K.P.2; Reiser M.F.1; Schoenberg S.O.1; 1. Institute of Clinical Radiology, University of Munich - Grosshadern Campus, Munich, Germany; 2. MR-Research, Bracco-Altana Pharma, Konstanz, Germany; 3. Department of Cardiovascular Radiology, UCLA, Los Angeles, CA.
Address correspondence to H.J. Michaely (henrik.michaely{at}med.uni-muenchen.de)
Objective: To evaluate the optimal technique for renal first pass perfusion imaging at 1.5T.
Materials and Methods: 15 healthy male volunteers (25-38 years) underwent in this prospective intraindividual comparison study measurements of renal perfusion on a 1.5T whole body MR-scanner (Magnetom Avanto, Siemens). Each volunteer was measured three times on three different days with three different sequences after the administration of 7ml Gd-BOPTA (Multihance®, Bracco-Altana Pharma) at 4ml/s: Saturation-recovery (SR) TurboFLASH without parallel imaging (TFl), SR-TrueFISP (TFi) and SR-Turbo-FLASH with parallel imaging (TFl-PI). The temporal resolution of the sequences was 4 slices/s (TFl, TFi) and 6 slices/s (TFl-PI). Field-of-view (450 x 373 mm2) and matrix (192 x 134) were kept constant for all sequences. Evaluation comprised the semiquantitative assessment of first pass perfusion parameters mean transit time (MTT), the time to peak signal intensity (TTP), the maximal upslope (MUS) and the maximal signal intensity (MSI). In addition the baseline signal-to-noise ratio (SNR) was assessed as well as the image quality which was rated by two radiologists in consensus.
Results: The MSI and hence the MUS were significantly lower with TFi (272 A.U./42 A.U./s) than with TFl (400 A.U./64 A.U./s) or TFl-PI (440 A.U./70 A.U./s). In contrast, the MTT (14.8s) and TTP (11.1s) were independent of the technique used. The image quality was best rated for TFl compared to TFi and TFl-PI. TFi revealed the highest baseline SNR, while TFl-PI exhibited the lowest SNR.
Conclusion: In conclusion, it seems that at 1.5T TFl without parallel imaging represents the best approach to renal first pass perfusion imaging.
275. Multidetector CT Urography (MDCTU) as a Screening Tool for Urinary Tract Neoplasms in Patients with Painless Hematuria
Sudakoff G.S.; Dunn D.P.*; Langenstroer P.; See W.A.; Eastwood D.C.; Dawson F.J.; Hellman R.S.; Radiology, Urology and Biostatistics, Medical College of Wisconsin, Milwaukee, WI.
Address correspondence to D.P. Dunn (ddunn{at}mcw.edu)
Objective: To determine if MDCTU is both sensitive and specific in detecting urinary tract neoplasms (UTN) in patients with painless hematuria.
Materials and Methods: During a 3-year-period, an electronic database was used to record the radiologic, urologic and pathologic records of 741 patients undergoing MDCTU. A retrospective review of this database was performed. MDCTU was performed using an identical protocol in all patients (non-infused phase, 2.5 mm collimation, 1.35 or 1.375:1 pitch followed by infused combined nephropyleographic phase, 1.25 mm collimation, 1.35 or 1.375:1 pitch). 475 of 741 patients were identified that underwent MDCTU and cystoscopy for painless hematuria without a prior history of (UTN). Urinalysis (UA) and urine cytology results were obtained in 340 and 352 of the 475 patients, respectively, who underwent MDCTU. The sensitivity and specificity of MDCTU in detecting UTN was evaluated. Multivariate logistic regression analysis was performed using the variables of: MDCTU diagnosis, worst urine cytology, UA (RBC 50 per hpf), age and gender to predict UTN.
Results: MDCTU correctly identified 36 of 39 patients with UTN (34 bladder, 4 pelvis, 1 ureter) for a sensitivity of 92%, specificity of 95.8%, positive predictive value of 70.4% and negative predictive value of 99%. Analysis with multivariate logistic regression found the variables of MDCTU diagnosis, worst urine cytology and UA (RBC > 50) to be significant (p < 0.0007). The odds for UTN in patients with abnormal MDCTU findings are 278 times greater then for those patients with normal MDCTU exams.
Conclusion: MDCTU is highly sensitive and specific in diagnosing UTN in patients without prior history of UTN and who have painless hematuria. In appropriate patients, MDCTU can be used as an effective screening tool for UTN.
276. Multi-Detector Row CT Angiography Evaluation of Living Related Laparoscopic Renal Donors: Is There a Need for Venous Phase Acquisition?
Namasivayam S.*; Kalra M.K.; Mittal P.; Torres W.E.; Small W.C.; Saini S.; Radiology, Emory University Hospital, Atlanta, GA.
Objective: To assess whether renal venous anatomy can be evaluated on the arterial phase MDCT alone.
Materials and Methods: IRB approval was obtained. Forty-eight renal donors (age range, 21-56 years; M: F, 20: 28) underwent MDCT angiography with 150 mL of 350 mg I/mL contrast material (4 mL/sec) on 8-slice (n = 40), or 64-slice (n = 8) MDCT. Arterial phase scan was triggered with bolus tracking and venous phase was started at delay of 55-60 seconds. Reconstructed slice thickness of 1.25-mm at 0.625-mm interval was used. Two radiologists reviewed both phases independently for renal venous anatomy, and graded venous enhancement on scale of 1-5 (1-not seen, 5-well opacified). Accuracy of arterial phase for venous anatomy was evaluated using venous phase as standard of reference. Venous enhancement for arterial and venous phases was compared. In addition, modified renal donor MDCT contrast injection protocol was used in five subjects to assess if optimum arterial and venous opacification is possible in single phase scanning.
Results: All venous anomalies noted on venous phase were also identified on arterial phase images. Accessory right renal vein (n = 14), retroaortic left renal vein (n = 2), circumaortic left renal vein (n = 1), enlarged left gonadal vein (n = 1), and left renal arteriovenous malformation (n = 1) were detected on arterial phase images by both reviewers. One reviewer recorded late right renal venous confluence (n = 2) in venous phase only, whereas, other reader did not detect this in both phases. Retroaortic left renal venous branch was reported as lumbar vein on arterial phase, and circumaortic left renal vein on venous phase (reviewer-1, n = 1; reviewer-2, n = 2). Venous phase images revealed significantly greater opacification of main renal veins, left gonadal, adrenal, and lumbar veins (p < 0.001). However, the opacification of the variant renal venous structures was not significantly different (p > 0.05). There was fair to good interobserver agreement (= 0.4-0.8). The new contrast injection protocol resulted in acceptable venous opacification and delineation of venous anatomy in a single phase scanning.
Conclusion: Arterial phase MDCT can detect renal venous anomalies. Though, opacification of left gonadal, adrenal and lumbar veins is limited in arterial phase, these can still be identified. Elimination of venous phase for evaluation of renal donors can reduce radiation dose and scan time. Moreover, our new protocol can enable optimum venous opacification in a single phase scanning.
277. Comparison of Distention and Opacification Utilizing Three-Phase vs. Split-Bolus Two-Phase Multi-Detector Row CT Urography
Dillman J.R.1; Caoili E.M.1; Cohan R.H.1; Ellis J.H.1*; Francis I.R.1; Nan B.1,2; 1. Radiology, University of Michigan, Ann Arbor, MI; 2. Public Health, University of Michigan, Ann Arbor, MI.
Address correspondence to J.H. Ellis (jimellis{at}umich.edu)
Objective: To compare urinary tract distention and opacification obtained with three-phase versus split-bolus two-phase multi-detector row CT urography (CTU).
Materials and Methods: Twenty-six consecutive normal three-phase (non-contrast, nephrographic, and excretory phase) CTU examinations were compared to 26 consecutive normal split-bolus two-phase (non-contrast and combined nephrographic/excretory phase) CTU studies independently utilizing two expert readers. Excretory phase images were reviewed and quantitative measurements of distension (axial images) and qualitative measurements of opacification (volume-rendered images) were recorded for predefined urinary tract segments. Images were also graded with respect to overall image quality, and split-bolus images were evaluated for the presence/degree of streak artifact.
Results: Three-phase CTU images provided superior renal collecting system distension (p = 0.04), opacification (p = 0.05), and image quality (p < 0.001) than did split-bolus CTU images. The three-phase protocol also demonstrated better bladder opacification (p < 0.001) and bladder image quality (p = 0.004). While there was greater ureteral distention, opacification, and image quality utilizing the three-phase technique, these differences were not statistically significant. Split-bolus streak artifact was typically absent (44%) or mild (46%) and only rarely moderate (10%).
Conclusion: Three-phase CTU produces better renal collecting system and urinary bladder opacification and image quality than does split-bolus two-phase CTU. Superior renal collecting system distention is also obtained with the three-phase technique. However, these improvements come at a cost of greater radiation exposure to the patient.
278. Characterization of Focal Renal Lesions with Low-mechanical Index Contrast-Enhanced Ultrasound versus CT and MRT
Clevert D.A.1*; Stickel M.2; Flach P.1; Strautz T.; Becker C.1; Schoenberg S.O.1; Reiser M.1; 1. Department of Clinical Radiology, University of Munich-Grosshadern, Munich, Germany; 2. Department of Surgery, University of Munich-Grosshadern, Munich, Germany.
Address correspondence to D.A. Clevert (Clevert{at}web.de)
Objective: To evaluate and compare the diagnostic accuracy of low -mechanical- index (low MI) contrast enhanced ultrasound (CUS) with MRT, CT and histological and operative findings in focal lesions of kidney.
Materials and Methods: 24 patients with focal renal lesions depicted by ultrasound were prospectively examined with CUS in low MI-technique using 1,6 to 2,4 ml intravenous injection of SonoVue (Bracco, Italy) using a multifrequency transducer 2-4 MHz (Logiq 9, General Electrics, Milwaukee, WI). The entire examination was saved as Cine-sequences. The contrast enhancement was examined in the arterial, venous, and late venous phase. The results were compared with CT or MRT. As gold standard histological findings were assessed.
Results: In all patients contrast enhanced US could be performed and perfusion pattern of the renal lesion could be diagnostically evaluated. Histological findings showed 16 solid renal cell carcinoma, 5 cystic lesions and 3 pseudo-tumors. Beside the different contrast patterns even smallest lesions could be depicted. In addition for pre-surgery planning the renal veins and inferior cava were examined.
Conclusion: CUS with SonoVue allows an early evaluation of smallest renal lesions. It is an additional examination to CT and MRT. Due to the dynamic examination additional information about tumor infiltration into the renal veins or inferior cava vein can be gained.
279. Is a Low Dose (80 kVp) Excretory Phase Scout Image a Viable Alternative to Thin Section CT-Urography for Accurate Delineation of Pyeloureteric Anatomy of a Kidney Donor?
Singh A.H.*; Sahani D.V.; Hahn P.F.; Mueller P.R.; Department of Abdominal Imaging, Massachusetts General Hospital, Boston, MA.
Address correspondence to A.H. Singh (dranandsingh{at}yahoo.com)
Objective: To study the value of a low kVp excretory phase (CT scout image as an alternative to thin section CT urography (CTU) for accurate delineation of pyeloureteric anatomy in living renal donors and its impact on the radiation dose.
Materials and Methods: Seventy renal donors were studied on a 16-MDCT scanner (GE light-speed). The arterial, venous and excretory phase (6 minutes delay) imaging was performed after intravenous injection of 125 cc of iodinated contrast media at 5 cc/sec. In 55/70 donors, a low kVp (kVp = 80) excretory phase (EP) scout images was also obtained prior to obtaining the CTU images (Distance covered/Speed = 2.5 mm/18mm, recons = 1.25 mm). Two blind readers to study the pyelouretric anatomy evaluated the low kVp EP-scout image and CTU images independently. Reader's confidence was graded on a 5-point scale. Findings at surgery constituted a standard of reference. The radiation dose delivered with each exam, number of images and time taken for CTU was recorded along with the findings.
Results: Normal bilateral pyelouretric anatomy was present in 52/55 patients and 3 patients demonstrated presence of ureteric duplication. There was optimal opacification of pyeloureteral system with 100% concordance of the EP-scout and CTU resultswith the surgical findings. Both the readers were comfortable with the low kVp scout image for interpretation (confidence scale = 4). The scan time was less for obtaining scout image (5s) compared to CTU (25s) with similar differences in the post processing time. The effective radiation dose to the patient dropped from 17 to 2 CT dose index (CTDI) with low kVp scout image.
Conclusion: A low kVp excretory phase scout image is effective in delineating the pyeloureteric anatomy and gross anomalies with similar effect as shown by CT urography and is advantageous over the latter in markedly reducing the effective radiation dose to the patient with simplified image processing.
280. CT Urography: Evaluation of Strategies for Opacification of the Collecting System
Sanyal R.*; Taori K.; Deshmukh A.; Department of Radiodiagnosis, Government Medical College, Nagpur, Maharashtra, India.
Address correspondence to R. Sanyal (rupansanyal{at}rediffmail.com)
Objective: To compare various protocols for the opacification of the collecting system in CT urography.
Materials and Methods: 60 patients underwent CT urography for various indications, during which 2.5 mm collimation scanning was done 8 to 10 minutes after contrast administration. In the time interval between the initial contrast administration and scanning in the pyelographic phase, various techniques were applied to maximize the opacification of the collecting system. Patients were scanned after 10mg iv furosemide injection (15 patients, 30 ureters); after 250 ml saline infusion in the supine position (15 patients, 30 ureters); after 250 ml saline infusion in prone position (15 patients, 30 ureters);and iv buscopan injection along with 250 ml saline infusion in the prone position (15 patients, 30 ureters). Buscopan was added due to its antiperistaltic effect on the ureters. During evaluation of the images, each collecting system and ureter was divided into six anatomic segments: upper and lower intrarenal collecting system, renal pelvis, and proximal (above iliac crest), lower (to the level of the sciatic notch), and distal ureter (below the sciatic notch). Each segment was assigned a score according to the opacification as follows: 0, unopacified; 1, less than 50% of segment opacified; 2, more than 50% of segment opacified; 3, entire segment opacified. Statistical analysis was done using paired Student t test to compare the opacification scores of the different acquisition techniques.
Results: The renal pelves and calyces were completely opacified in all four groups. Furosemide administration resulted in complete opacification of 93% ureters (28 of 30). In distal ureter, the mean score with furosemide was 2.9 while that with saline supine and prone positioning was 1.86 and 1.83 respectively, and this difference was highly significant (p = 0.0001). It was also significantly higher than the buscopan group (score 2.3, p = 0.0154). In mid and upper ureter also furosemide had significantly higher scores than saline infusion in either position. Saline infusion in supine and prone positions had very similar scores in all segments which were less than the buscopan group but this difference was not statistically significant.
Conclusion: During CT urography, furosemide administration in very low doses is highly effective in significantly improving the delineation of the ureters as compared to saline infusion in either prone or supine position, or addition of buscopan to saline supplementation.
281. Hepatorenal Syndrome: Quantitative Evaluation with Contrast-Enhanced MRI
Woo D.H.*; Benstein J.A.; Lee V.; Hecht E.; Taouli B.; Radiology, NYU Medical Center, New York, NY.
Address correspondence to D.H. Woo (wood01{at}endeavor.med.nyu.edu)
Objective: To quantify renal cortical and medullary enhancement in cirrhotic patients with hepatorenal syndrome (HRS), using gradient echo T1-weighted imaging before and after dynamic injection of gadolinium contrast.
Materials and Methods: Retrospectively, we identified 13 patients with cirrhosis and acute renal dysfunction [serum creatinine (SCr) level > 2.0 mg/dl] who underwent MRI with a presumed diagnosis of HRS (9 men, mean age, 59 y). Patients with chronic renal dysfunction were excluded. SCr was measured with a mean delay of 2 days (range 0-7 days) of MRI. All patients had advanced cirrhosis (Child-Pugh B or C, n = 3 and n = 10, respectively). The average SCr level was 4 mg/dl (range, 2-8). We also included a control population of 10 patients with cirrhosis and normal SCr (6 men; mean age, 55 y). MRI of the liver, including images of at least a portion of the kidneys, was obtained at 1.5 T using the following sequences: coronal HASTE, axial T1 in- and out-of-phase GRE, axial T2 STIR, axial 3D T1-weighted GRE (VIBE) or, for poor breath-holders, 2D turboFLASH, before and after gadolinium contrast injection [arterial (Art), portal venous (PV), and delayed (Del) phases], with the first acquisition timed using a test bolus and subsequent acquisitions at 45 sec and 120 sec). Quantitative analysis of renal enhancement was performed by measuring signal intensity (SI) values using ROIs positioned over renal cortex and medulla on pre- and post-contrast images. The relative enhancement (Rel Enh) of renal cortex and medulla was calculated for all three phases, [(SIpost-SIpre)/SIpre]x100%, and values for HRS compared with control patients.
Results: Relative enhancement was significantly lower for HRS patients than controls for all measurements except for the medullary arterial phase. Measurements for patients with HRS vs. patients with normal renal function were: Cortex-ART Rel Enh 111% vs. 254%, Cortex-PV Rel Enh 130% vs. 273%, Cortex-Del Rel Enh 113% vs. 256%; Medulla-PV Rel Enh 153% vs. 320%, Medulla-Del Rel Enh 156% vs. 362% (p < 0.001 for all ratios).
Conclusion: Patients with HRS demonstrated significantly decreased quantitative cortical and medullary enhancement at all phases (except for medulla at the arterial phase). Contrast enhanced MRI may potentially be used prospectively for earlier detection and treatment of HRS.
282. Value of 3D Reconstructions in the Evaluation of Genitourinary Tract with 16- and 64-Slice MDCT
Sebastian S.*; Kalra M.K.; Bernardino M.E.; Baumgarten D.A.; Radiology, Emory University Hospital, Atlanta, GA.
Address correspondence to S. Sebastian (Sunit.Sebastian{at}emoryhealthcare.org)
Objective: Most radiologists use only transverse images for interpretation of abdominal CT for evaluation of genitourinary tract. The purpose of our study was to determine if review of batch coronal multiplanar reformats (coronal MPR) supplements information obtained from transverse images of abdominal-pelvic CT performed for genitourinary indications.
Materials and Methods: IRB approval was obtained. In an on going study, we have reviewed 50 abdominal-pelvic MDCT studies (64 slice MDCT, n = 42; 16-slice, n = 8), performed for genitourinary tract evaluation in 50 patients. Scanning parameters were held constant for both scanners. Transverse images were reconstructed at 5 mm and 0.625 mm. Thin transverse images were used for generating batch coronal MPR with 3 mm thickness. Prior to the study, two abdominal radiologists underwent a training session for interpretation of coronal MPR on 10 separate studies. Subsequently, each reader reviewed transverse images of all the patients in a blinded fashion. After 3 weeks, both transverse and coronal MPR images were reviewed. For each review, both radiologists recorded the number of lesions, their location, size of smallest lesion, probable diagnosis and confidence for interpretation (3 point scale, 1 = not confident, 3 = certain). Image quality (5 point scale, 1 = unacceptable, 5 = excellent) was also assessed. Statistical analysis was performed using paired t-test.
Results: Both readers reported significantly more findings on combined review of transverse and coronal images (226 and 232) compared to transverse images alone (211 and 215) (p < 0.03). Most additional findings on combined review of images were vascular anomalies (n = 18). Combined review of images helped in excluding two pseudolesions (pseudotumor and bladder calculus). Both readers had significantly greater confidence for combined review of transverse and coronal images when compared to transverse images alone (p < 0.0001). Image quality of both axial and coronal images was acceptable (average image quality scores greater than 3.6 for both readers).
Conclusion: 16- and 64-slice MDCT scanners provide acceptable quality of coronal MPR images for evaluation of genitourinary tract. Initial results suggest that addition of multiplanar coronal MPR to transverse images can provide important supplementary information for evaluation of genitourinary tract with MDCT scanning.
283. Angiomyolipoma with Minimal Fat on MDCT: Can Pixel Counts Aid Diagnosis?
Simpfendorfer C.S.1*; Herts B.R.1; Remer E.M.1; Magi-Galluzzii C.2; Zhou M.2; Lockwood D.1; Lieber M.1; Motta-Ramirez A.G.1; 1. Diagnostic Radiology, Cleveland Clinic Foundation, Cleveland, OH; 2. Anatomic Pathology, Cleveland Clinic Foundation, Cleveland, OH.
Address correspondence to C.S. Simpfendorfer (simpfec3{at}ccf.org)
Objective: To determine the positive predictive value and sensitivity of subzero pixel value counts for the diagnosis of renal AML and correlate findings with pathologic assessment of fat content.
Materials and Methods: 33 AML were identified from 719 renal mass resections performed in 702 patients over 3 years. Of 33 AML, 15 AML were prospectively diagnosed based on fat at MDCT. The remaining 18 not diagnosed prospectively were included in the study. A cohort of renal cell carcinomas matched for age, gender and tumor size was identified from the larger group. 3 radiologists blinded to the diagnosis reported the visual presence of fat and then measured the number of pixels less then -10, less then -20, and less then -30 HU in all lesions. The sensitivity, specificity, and positive predictive value adjusted for prevalence were calculated using 3 different criteria based on the ROC curves; (i) more then 10 pixels less then -20 HU, (ii) more then 20 pixels less then -20 HU, and (iii) more then 5 pixels less then -30 HU. Two pathologists retrospectively reported the percentage of fat in 14 AML as more or less than 10%. The presence of negative pixels was correlated with pathologic determination of fat content.
Results: On visual inspection reader A identified 3, reader B 5, and reader C 4 of the 18 AML. The combined sensitivity was 22%, specificity 98%, and PPV 72%. Using criteria (i), reader A identified 6, reader B 5, and reader C 2 AML. The combined sensitivity was 24%, specificity 98%, and PPV 69%. Using criteria (ii), Reader A identified 3, Reader B 4, and Reader C 2 AML. The combined sensitivity was 17%, specificity 100%, and PPV 100%. Using criteria (iii), Reader A identified 4, Reader B 4, and Reader C 2 AML. The combined sensitivity was 18%, specificity 100%, and a PPV 100%. 8 of 14 AML (43%) contained less then 10% fat at pathology; all 8 had 0 pixels less then -20 HU. There is a statistically significant correlation between AML with no negative attenuation pixels at CT and AML with less than 10% fat at pathology.
Conclusion: CT findings of more than 20 pixels less then -20 HU and more then 5 pixels less then -30 HU have a positive predictive value of 100%, but many AML with minimal fat cannot be diagnosed preoperatively by pixel counts. AML with MDCT measurements demonstrating no fat have minimal fat at pathology, and cannot reliably be diagnosed preoperatively.
284. Comparison of CT Histogram Analysis to Adrenal Washout CT in the Diagnosis of Lipid Poor Adenomas
Jhaveri K.S.*; Lad S.V.; Haider M.A.; Medical Imaging, University Health Network and Mount Sinai Hospital, Toronto, ON, Canada.
Address correspondence to K.S. Jhaveri (kartik.jhaveri{at}uhn.on.ca)
Objective: CT histogram analysis has been proposed as a simple post processing technique to diagnose an adrenal adenoma by demonstration of significant intralesional negative pixels (typically greater than 5% or 10%). The purpose of this retrospective pilot study was to assess the potential of CT histogram analysis in diagnosing an adrenal adenoma within the lipid poor category in comparison to adrenal washout CT.
Materials and Methods: Ten patients (age range: 51-74 years, 6 females and 4 males) with twelve lipid poor adrenal adenomas (size range: 13-42 mm) were identified from the radiology reporting database. Lipid poor adenoma was defined as a nodule having unenhanced CT attenuation more than 10 HU but showing 60% or greater absolute enhancement washout on adrenal washout CT. Adrenal washout CT had been performed in all cases by obtaining 2.5mm CT slices on a multislice CT scanner during unehanced, 60-second post contrast and 15 minute delayed scans. Histogram analysis was performed as post processing on a PACS workstation on the unenhanced CT phase. A region of interest (ROI) was drawn within the nodule recording mean attenuation, number of pixels and percentage of negative pixels. A diagnosis of adenoma was made if there was a greater than 5% negative pixel threshold on the histogram analysis.
Results: The mean attenuation in all nodules was > 10 HU (range 12-30 HU) and absolute enhancement % washout was 60 or greater (60-79%) indicating a diagnosis of lipid poor adenoma by adrenal washout CT. CT Histogram analysis revealed negative pixels in all 12 adenomas. The negative pixel % ranged from 3% to 20%. Of the 12 adenomas, 11 (91.6%) showed > 5% negative pixels and 7 of 12 (58.3%) showed > 10% negative pixels There was good correlation between mean attenuation and % negative pixels (Pearson's product-moment correlation = -0.85, 95% confidence interval: -0.96, -0.52, p-value = 0.0008) and there was no significant correlation between the absolute enhancement % washout and the % negative pixel values. (Pearson's product-moment correlation = -0.39, 95 % confidence interval: -0.8, 0.28, p-value = 0.2406)
Conclusion: The preliminary results are encouraging for the application of CT histogram analysis in characterization of lipid poor adrenal adenomas. If this can be confirmed in larger sample size studies there is a potential for reducing the actual number of adrenal washout CT examinations for diagnosis of lipid poor adenomas.
* Will present paper
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
