Hepatic MR imaging: comparison of RARE derived sequences with conventional sequences for detection and characterization of focal liver lesions

Background: We compared two T2-weighted turbo spin echo (TSE) sequences with a T2-weighted conventional SE (CSE) sequence to determine whether sequences derived from rapid acquisition with relaxation enhancement such as TSE could replace CSE for the detection and subsequent characterization of focal liver lesions. Methods: A total of 55 consecutive patients with 107 liver lesions underwent magnetic resonance imaging examinations at 1.5 Tesla, with a constant imaging protocol. TSE pulse sequences were acquired with eight echo trains (repetition time [TR], 4718 ms; echo time [TE], 90 ms; acquisition time [TA], 4.03 min; and a symmetric k-space ordering scheme) and 11 echo trains (TR, 4200 ms; TE, 140 ms; TA, 4.40 min; and an asymmetric k-space ordering scheme) and compared with CSE (TR, 2300 ms; TE, 45/90 ms; TA, 9.53 min). Images were analyzed qualitatively by scoring image quality and artifacts and counting focal liver lesions by independent reading with consensus obtained for discrepancies. Quantitative analysis was performed by measuring signal-to-noise (S/N), contrast-to-noise (C/N), and tumor–liver signal intensity (T/L) ratios. Results: T2-weighted TSE sequences provided better subjective image quality and reduced artifacts as compared with the T2-weighted CSE sequence. CSE and TSE sequences exhibited no statistically significant differences in liver S/N, lesion–liver C/N (CSE TE, 90 ms: 18.6 ± 14.0; TSE TE, 90 ms: 16.5 ± 12.9) and the detectability of focal liver lesions. Heavily T2-weighted TSE with a TE of 140 ms allowed correct characterization of focal liver lesions based on a T/L ratio of 3.0 in 84% of patients. Conclusions: T2-weighted TSE sequences are as suited as CSE for the detection (TE, 90 ms), and appear to be superior for the characterization (TE, 140 ms), of focal hepatic lesions. Whether a single sequence, such as a double-echo TSE or a single-echo TSE sequence with a TE between 110 and 120 ms, might perform both functions as well or better than CSE is unknown. However, because of time savings, TSE eventually may be preferred over CSE. Received: 13 December 1994/Accepted after revision: 31 March 1995     

Small hepatocellular carcinoma: differentiation from adenomatous hyperplasia with color Doppler US and dynamic Gd-DTPA-enhanced MR imaging

Background: To investigate the usefulness of color Doppler ultrasound (US) and dynamic Gd-DTPA-enhanced magnetic resonance (MR) imaging in the differentiation of small hepatocellular carcinoma (HCC) and adenomatous hyperplasia (AH). Methods: Thirty-eight small (3 cm or less) nodular lesions (in 38 cirrhotic patients) with US features consistent with HCC underwent evaluation with color Doppler US and MR imaging. Breath-hold T1-weighted rapid acquisition spin echo MR sequence after bolus injection of 0.1 mmol/kg gadopentetate dimeglumine was used to evaluate dynamic enhancement. US-guided tissue-core percutaneous biopsy established the diagnosis: HCC in 28 cases and AH in 10. Results: Color signals with pulsatile or continuous Doppler spectrum were demonstrated in 19 of 28 HCCs (68%) but in none of the AHs. Although there was considerable overlap in signal intensity between HCC and AH on both unenhanced T1- and T2-weighted images, early enhancement on breath-hold T1-weighted images obtained 40 s after starting contrast administration was observed in 22 of 28 HCCs (79%) but in none of the AHs. In 26 of 28 HCCs (93%), pulsatile or continuous flow at color Doppler US, early enhancement at dynamic MR imaging, or both were observed. Conclusion: Findings with color Doppler US and dynamic Gd-DTPA-enhanced MR imaging enable a reliable distinction between small HCC and AH. Received: 8 August 1994/Accepted after revision: 27 January 1995     

Plain and gadolinium-DTPA-enhanced MR imaging of hepatocellular carcinoma treated with transarterial chemoembolization

Background: To assess unenhanced and gadolinium-enhanced magnetic resonance (MR) imaging patterns of hepatocellular carcinoma (HCC) treated with transarterial chemoembolization (TACE). Methods: Thirty-two patients with 48 HCC lesions underwent MR imaging before and 15 days after TACE. Fifteen lesions were then surgically resected. The remaining 33 lesions were not removed and were followed up with MR imaging at 3, 6, 12, and 18 months after treatment. Spin echo (SE) T1- and T2-weighted and gadolinium-enhanced SE T1-weighted sequences were employed. Qualitative evaluation of signal intensity pattern of the treated lesions was performed in all cases. Histological evaluation and selective hepatic arteriography were considered the gold standard of the study for the 15 resected lesions and the 33 unresected lesions, respectively. Results: On follow-up enhanced T1-weighted images of the 15 resected lesions, seven showed no area of enhancement corresponding to complete necrosis at histologic examination. The remaining eight resected lesions showed areas of enhancement; in six of these cases, viable tumor tissue was found at histology; in the other two lesions, histologic examination revealed the presence of complete tumor necrosis. In the group of resected lesions, T2-weighted images showed no pattern characteristic of necrosis. In 24 of 33 unresected lesions, loss of enhancement on follow-up enhanced T1-weighted images was a characteristic finding, which correlated to devascularization at arteriography. Of these 24 lesions, 17 were completely hypointense on follow-up T2-weighted images; the remaining seven showed small foci of hyperintensity. The other nine unresected lesions showed enhanced portions on follow-up enhanced T1-weighted images, which corresponded to hyperintense areas on T2-weighted images. These findings correlated to persistence of hypervascular areas at arteriography. Conclusion: Gadolinium-enhanced T1-weighted MR imaging is a reliable method for evaluating the outcome of TACE treatment and is more accurate than unenhanced T2-weighted MR imaging. Received: 2 June 1995/Accepted: 18 July 1995     

Diffusion-weighted MR imaging with single-shot echo-planar imaging in the upper abdomen: preliminary clinical experience in 61 patients

Background: To determine the potential ability of diffusion-weighted magnetic resonance (MR) imaging with single-shot echo-planar imaging (DW imaging) in the upper abdomen by apparent diffusion coefficient (ADC) and signal:intensity ratio (SIR) measurements. Methods: DW imaging was performed in 61 clinical patients. ADCs in the liver, pancreas, spleen, kidney, and different pathological conditions were calculated. Spleen-to-liver SIR and segmental intensity difference of the liver (SID) were also calculated. Results: The mean ADCs (mm²/s) were 2.28 × 10⁻³± 0.07 in the liver, 1.44 × 10⁻³± 0.05 in the spleen, 1.94 × 10⁻³± 0.19 in the pancreas, and 5.76 × 10⁻³± 0.06 in the kidney. The mean ADC of cirrhotic liver was 1.96 × 10⁻³± 0.62, which was lower than that of normal liver. Other pathologic conditions also showed ADCs different from those of normal tissues. All DW images showed significantly higher spleen-to-liver SIRs and SIDs than did T2-weighted images (p < 0.05). Conclusion: The mean ADCs obtained with DW imaging were different in each upper abdominal organ and with each pathologic condition. DW images showed better soft tissue contrast than did T2-weighted images with regard to SIR and CNR in depicting and characterizing upper abdominal disorders. Received: 24 July 1998/Revision accepted: 2 December 1998     

MR imaging of advanced gastric cancer: comparison of various MR pulse sequences using water and gadopentetate dimeglumine as oral contrast agents

Background: To evaluate clinical usefulness of oral contrast agents (gadopentetate dimeglumine and water) and to assess proper magnetic resonance (MR) imaging in evaluating advanced gastric cancer (AGC) by comparing different MR imaging techniques. Methods: Fifteen patients with AGC were imaged with a 1.0-T MR imager and body-array coil. All patients underwent surgery or laparascopic biopsy. Fast low-angle shot (FLASH), half-Fourier single-shot turbo spin-echo (HASTE), and true fast imaging with steady-state precession time (FISP) images were obtained after ingestion of 900 mL tap water in each patient, followed by postcontrast FLASH images after additional ingestion of gadopentetate dimeglumine (Gd-DTPA). Qualitative analysis including T-staging of AGC and scoring of imaging quality and quantitative analysis were performed prospectively. Results: In image quality and diagnostic accuracy of T-staging, FLASH imaging showed results slightly superior to those of other imaging modalities, and there was no great difference between using water and Gd-DTPA as an oral contrast agent. As for cancer-to-gastric lumen contrast-to-noise ratio (CNR), HASTE and true FISP imaging were superior to FLASH imaging with Gd-DTPA (p < 0.0001). In cancer-to-pancreas CNR, FLASH imaging without Gd-DTPA showed the best result. Conclusions: The use of Gd-DTPA as a positive contrast agent may not be imperative, and T1-weighted FLASH imaging in combination with true FISP imaging with ingestion of tap water can be very useful in evaluating AGC with MR imaging. Received: 29 October 1998/Revision accepted: 27 January 1999     

Dynamic MR imaging of hepatolithiasis

Background: To report the dynamic magnetic resonance (MR) imaging findings of hepatolithiasis. Methods: Dynamic MR images (fast spoiled gradient echo sequence with intravenous injection of gadopentate dimeglumine) and computed tomography, cholangiography, or angiography of nine patients with hepatolithiasis are analyzed. Results: All affected hepatic segments showed atrophic changes and contained dilated intrahepatic ducts. These segments showed either iso- or hypointensity on T1-weighted imaging and hyperintensity on T2-weighted imaging. Preferential enhancement was noted throughout all phases of the dynamic study and persisted to delayed T1-weighted imaging in seven patients. In the last two patients, severe atrophic changes made evaluating signal intensity differences and enhancement patterns difficult. Conclusion: In addition to intrahepatic stones and biliary dilatation, segmental atrophy, signal intensity differences, and preferential and persistent enhancement are important MR findings of hepatolithiasis. Received: 18 June 1997/Accepted: 23 July 1997     

Focal nodular hyperplasia of the liver: detection and characterization with plain and dynamic-enhanced MRI

Background: We compared nonenhanced and dynamic gadolinium (Gd)–enhanced magnetic resonance imaging (MRI) appearances of hepatic focal nodular hyperplasia (FNH) as depicted with breath-hold MR sequences and assessed the detectability of the individual MR sequences used. Methods: We retrospectively reviewed 48 consecutive patients with FNH. All patients underwent nonenhanced (T1 fast low-angle shot [FLASH] and T2 half-Fourier acquisition [HASTE]) and dynamic Gd-enhanced (T1 FLASH) MRI between December 1997 and March 2000. Individual MR sequences were analyzed separately for number of lesions, signal intensity features, dynamic enhancement pattern, and the presence and enhancement profile of a central scar. Ninety-five percent confidence intervals of absolute discrepancy were calculated to define differences in lesion detection. Results: Seventy-seven lesions were found in 48 patients. Nonenhanced FLASH imaging depicted 59 (76.6%) lesions in 45 patients. HASTE images showed 55 (71.4%) lesions in 44 patients. On T1- and T2-weighted images, lesions appeared predominantly hypointense (69.5%) and hyperintense (72.7%), respectively. Arterial and portal venous dominant phase Gd-enhanced MRI demonstrated all 77 lesions (100%), most of which showed hypervascular (94.8%), homogeneous (97.4%), and incomplete (except the central scar: 58.4%) enhancement in the arterial phase. Portal venous phase images showed lesion isointensity (50.6%) or moderate hyperintensity (46.8%) with complete enhancement (central scar: 94.8%). A central scar was detected on nonenhanced T1-weighted images (hypointense: 100%), T2-weighted images (hyperintense: 100%), arterial phase (hypointense: 59.7%) and portal venous phase (hyperintense: 71.4%) Gd-enhanced images in 78%, 69.1%, 77.9%, and 75.3% of tumors, respectively. Conclusion: Arterial and portal venous phase Gd-enhanced T1-weighted sequences are superior to nonenhanced images in the detection of FNH. Typical MRI appearances include hypointensity on T1-weighted and hyperintensity on nonenhanced T2-weighted images. Most commonly, FNH shows a homogeneous (without scar) and strong enhancement during the arterial phase, with lesion isointensity or slight hyperintensity during the portal venous phase. Received: 15 May 2001/Revision accepted: 22 August 2001     

Well-differentiated hepatocellular carcinoma: findings of US, CT, and MR imaging

Background: To elucidate the imaging characteristics of well-differentiated hepatocellular carcinomas (HCCs) on ultrasonography (US), computed tomography (CT), and magnetic resonance (MR) imaging. Methods: Ultrasonograms, CTs, and MR images of 18 histopathologically proven well-differentiated HCCs in 15 patients were reviewed. The findings of these images were correlated with histopathologic findings. Results: On US, seven tumors were depicted as a hyperechoic area and eight as a hypoechoic area. Three tumors were not visualized. On precontrast CT, four tumors were depicted as a low-density area, but 14 were not visualized. On conventional contrast-enhanced CT, 12 tumors were depicted as a low-density area but six were not visualized. On T1-weighted MR images, 10 tumors had high signal intensity and two had low signal intensity. Six tumors were not visualized. On T2-weighted MR images, five tumors had high signal intensity and two had low signal intensity. Eleven tumors were not visualized. Tumors with fatty change and/or clear cell formation were frequently hyperechoic on US and hyperintense on T1-weighted MR images. Conclusions: Well-differentiated HCCs show different findings on US, CT, and MR imaging. Therefore, reliable diagnosis of well-differentiated HCCs by these imaging techniques may be difficult. Received: 29 April 1998/Revision accepted: 15 July 1998     

Fast T2-weighted liver MR imaging: comparison among breath-hold turbo-spin-echo, HASTE, and inversion recovery (IR) HASTE sequences

Background: To evaluate the diagnostic efficacy of fast T2-weighted magnetic resonance (MR) imaging sequences on image quality, hepatic lesion detection, and lesion conspicuity. Methods: Three breath-hold, fast T2-weighted sequences with turbo-spin-echo (TSE), half-Fourier acquisition single-shot TSE (HASTE), and inversion recovery (IR) HASTE techniques were examined for 43 lesions in 20 consecutive patients. Evaluation was performed qualitatively on image quality and lesion detectability and quantitatively on lesion conspicuity by using lesion/liver signal-intensity and contrast-to-noise ratios. Results: Artifacts were significantly less present on the HASTE sequence (p < 0.01). Both TSE and HASTE sequences detected 39 lesions (91% each); the IR HASTE sequence detected 37 (86%). IR HASTE sequence showed a significantly higher signal-intensity ratio than did the others (p < 0.01). Conclusions: Breath-hold TSE versus breath-hold HASTE or IR HASTE is still the most robust sequence in lesion detection, image quality, and lesion conspicuity. However, the HASTE sequence offers good lesion detection and image quality, and the IR HASTE has a better signal-intensity ratio. Received: 15 January 1999/Accepted: 24 March 1999     

MR imaging of cervical carcinoma: comparison among T2-weighted,dynamic, and postcontrast T1-weighted images with histopathological correlation

Background: To identify the reasons for misdiagnosis of the degree of stromal invasion by uterine cervical cancer with various magnetic resonance sequences. Methods: T2-weighted, dynamic, and postcontrast T1-weighted images were obtained in the sagittal plane in 20 patients with uterine cervical cancer. After evaluating these sequences for the degree of stromal invasion, histologic specimens were directly correlated with these images. Results: The degree of stromal invasion was correctly diagnosed in 15 of the 20 cases on T2-weighted images, in 12 on dynamic images, and in eight on postcontrast T1-weighted images. All misdiagnoses were due to overestimation. Histologically, peritumoral stroma showed inflammation or edema in two patients, whereas no histological abnormality was found in the other patients. A hyperintense rim, i.e., a peritumoral enhanced ring-shaped structure, was observed on the enhanced images of five patients. The hyperintense rim corresponded to the periphery of the tumor in three patients and to the cervical stroma in two patients. Conclusion: T2-weighted images permitted the most accurate evaluation of stromal invasion by uterine tumors. Overdiagnosis may be due to an abnormal intensity of the cervical stroma, which was observed more frequently on dynamic and postcontrast T1-weighted images than on T2-weighted images. Received: 10 November 1995/Accepted after revision: 13 March 1996     

Hepatic iron deposition on MR imaging in patients with chronic liver disease: correlation with serial serum ferritin concentration

Background: To evaluate the relationship between magnetic resonance (MR) imaging grading of iron deposition and serial serum ferritin concentration in patients with chronic viral liver diseases. Methods: In 80 patients with viral hepatitis and cirrhosis, MR images including T2*-weighted gradient echo images (echo time ≥ 6.5 ms) were reviewed. The grades of parenchymal iron deposition and iron-containing nodules in the liver and spleen and the liver-to-muscle and spleen-to-muscle signal intensity ratios were compared with the most recent, the mean, the lowest, and the highest values from all available serum ferritin levels. Results: The serum ferritin concentration was significantly correlated with the grades of iron deposition in liver and spleen and with the grades of iron-containing nodules seen on MR images (p < 0.05). Liver-to-muscle signal intensity ratio was weakly correlated with the ferritin concentrations. Among categories of ferritin concentration, correlation with MR grades was highest for mean ferritin concentration (r = 0.487, p < 0.001). Conclusion: MR imaging grades of hepatic iron and siderotic nodules correlate with serum ferritin, especially with the mean levels. Received: 9 May 2000/Accepted: 28 June 2000     

Dysplastic nodules of the liver: imaging findings

Background: To verify characteristic features of hepatic dysplastic nodules at different imaging modalities. Methods: Twenty-eight patients with 37 dysplastic nodules of the liver (0.8–3.0 cm) underwent sonography (28 patients), computed tomography (CT; 24 patients), magnetic resonance (MR; 11 patients), and angiography (12 patients). Each nodule was analyzed for echogenicity, attenuation, signal intensity, and vascularity. Results: Echogenicity of nodules was high in 16 (43%), homogeneous in two (6%), and low in 19 (51%) of 37 nodules. Attenuation of nodules was high in one (7%), homogeneous in four (26%), and low in 10 (67%) of 15 nodules on the arterial-phase CT images; homogeneous in five (33%) and low in 10 (67%) of 15 nodules on the portal-phase CT images; and high in four (17%), homogeneous in six (26%), and low in 13 (57%) of 23 nodules on the delayed-phase CT images. Signal intensity of nodules was high in 15 (94%) and homogeneous in one (6%) of 16 nodules on T1-weighted MR images and was homogeneous in seven (44%) and low in nine (56%) of 16 nodules on T2-weighted MR images. Vascularity of nodules was avascular in 14 (88%) and slightly vascular in two (12%) of 16 nodules. Conclusions: Hepatic dysplastic nodules show diverse imaging characteristics with different imaging techniques; however, common imaging findings of hepatic dysplastic nodules are low echo, low attenuation, and high, low, or homogeneous intensity on T1- and T2-weighted MR, and avascularity. Received: 13 May 1998/Accepted: 1 July 1998     

Detection of lymph-node metastases in patients with gastric carcinoma: comparison of three MR imaging pulse sequences

Background: To compare the diagnostic accuracy of magnetic resonance (MR) images obtained with three different pulse sequences for lymph-node metastases in patients with gastric cancer. Methods: T1-weighted spin-echo (SE), breath-hold T2-weighted fast SE, and triphasic gadolinium-enhanced dynamic gradient-recall-echo (GRE) MR images obtained in 16 patients with gastric carcinoma were retrospectively reviewed. Regional lymph nodes were assigned to four different groups, and image review was conducted on a lymph-node group-by-group basis; 64 lymph-node groups were reviewed by two radiologists. Relative sensitivity, specificity, and accuracy were determined based on the findings with definitive surgery and follow-up imaging. Diagnostic accuracy was determined by means of receiver-operating-characteristic (ROC) analysis. Results: Relative sensitivities for lymph-node metastases with T1-weighted SE, breath-hold T2-weighted fast SE, and dynamic GRE images were 61%, 94%, and 59%, respectively. Relative sensitivity with breath-hold T2-weighted fast SE images was significantly greater than that with T1-weighted SE (p < 0.05) and dynamic GRE (p < 0.05) images. Diagnostic accuracy determined by ROC analysis was marginally higher with breath-hold T2-weighted fast SE (area under ROC curve [Az]= 0.87) than with T1-weighted SE (Az = 0.78, p= 0.08) and dynamic GRE (Az = 0.79, p= 0.12) images. Conclusion: Breath-hold T2-weighted fast SE sequence is useful in the detection of regional lymph-node metastases in patients with gastric carcinoma. Received: 11 November 1998/Revision accepted: 7 April 1999     

T2-weighted MR imaging for hepatic hemangiomas: comparison of breath-hold and non-breath-hold turbo spin-echo pulse sequences with phased-array multicoil

Background: We compared T2-weighted and heavily T2-weighted breath-hold turbo spin-echo (TSE) sequences with T2-weighted non-breath-hold TSE sequence to evaluate hepatic hemangiomas on magnetic resonance (MR) with a phased-array multicoil. Methods: Twenty-two patients with 27 hemangiomas were studied at 1.0-T scanner by using T2-weighted and heavily T2-weighted breath-hold TSE sequences (18 s each) and non-breath-hold T2-weighted TSE sequences with use of a phased-array multicoil. Images were quantitatively analyzed for tumor-to-liver signal-difference-to-noise ratios (SD/Ns) and tumor-to-liver signal intensity ratios (T/Ls) and qualitatively analyzed for tumor conspicuity and motion-induced image artifacts. Results: Quantitatively, T2-weighted breath-hold TSE images showed the highest SD/Ns among the three sequences, although the differences from the heavily T2-weighted breath-hold TSE sequence and the T2-weighted non-breath-hold TSE sequence were not statistically significant (p= 0.61 and 0.06, respectively). Heavily T2-weighted breath-hold TSE images showed the highest T/Ls among the three sequences. The differences from the T2-weighted breath-hold TSE sequence and the T2-weighted non-breath-hold TSE sequence were statistically significant (p < 0.001). Qualitatively, breath-hold TSE images were superior to non-breath-hold TSE images in terms of tumor conspicuity (p < 0.01) and motion artifacts (p < 0.01). Conclusion: T2-weighted breath-hold TSE sequence is superior to T2-weighted non-breath-hold TSE sequence in the evaluation of hepatic hemangiomas on MR with a phased-array multicoil. Received: 7 April 1997/Accepted: 28 May 1997     

肝脏实性占位磁共振成像定量指标与病灶性质的相关研究

目的:研究病灶?肝脏信号强度比(SIR)、信噪比(SNR)、对比噪声比(CNR)在区分肝脏实质性小占位的性质方面的价值。方法:随机选择经确诊的肝细胞癌、肝转移瘤和肝血管瘤共67例(101个病灶),行前瞻性磁共振成像(0.5T)研究。计算病灶的SIR,SNR,CNR,并进行统计学分析。结果:T1W图像及质子加权像上,三种定量指标对区分肝血管瘤和肝细胞癌、肝转移瘤无价值。T2W图像上,肝转移瘤与肝细胞癌的SIR,SNR,CNR值均显著小于血管瘤的SIR,SNR,CNR值,TE=60ms时SIR,SNR,CNR区分肝血管瘤和肝细胞癌的诊断正确率分别为72%,72%,83%;TE=90ms时诊断正确率分别为74%,80%,84%;TE=120ms时诊断正确率分别为75%,86%,88%。结论:T1W图像及质子加权像上,三种定量指标对区分肝脏实质性占位的性质无明确价值,T2W图像上测得的SIR,SNR,CNR值可用于区分肝血管瘤与肝细胞癌,且TE=120ms时诊断正确率最高。     

Intestinal Behçet's disease: breath-hold MR imaging

Background: To determine the feasibility of magnetic resonance (MR) imaging in the evaluation of intestinal Behçet's disease. Methods: Eight patients diagnosed to have intestinal Behçet's disease prospectively underwent MR imaging. Five patients had previously undergone abdominal surgery for intestinal Behçet's disease 27–81 months previously. Breath-hold T2-weighted single-shot fast spin echo, T1-weighted multiplanar spoiled gradient recalled echo, and postgadolinium-enhanced fat-suppressed T1-weighted opposed-phase multiplanar spoiled gradient recalled echo images were obtained. MR findings were analyzed by three independent radiologists in terms of the presence and location of bowel wall thickening and ulcer, maximal bowel wall thickness, grade of bowel wall enhancement, and presence and nature of extraluminal manifestations. K statistics were used to evaluate interobserver variability. Results: Seven cases (88%), including the five cases with previous operations, demonstrated bowel thickening of more than 9 mm and increased enhancement. Ulcer was depicted in four cases (50%) in the terminal ileum or an anastomosis site. Three cases (38%) showed extraluminal manifestation, such as mesenteric infiltration around the involved bowel (two cases) and a sinus tract to subhepatic abscess from perforation (one case). Conclusion: Breath-hold MR imaging is useful for the evaluation of bowel wall thickening, enhancement, mesentery infiltration, and abscess formation. Characteristic ulcerative lesions may be also depicted on T1- and T2-weighted breath-hold images. MR imaging is a useful method for postoperative follow-up study for the evaluation of recurrence.     

Liver MR imaging: comparison of respiratory triggered fast spin echo with T2-weighted spin-echo and inversion recovery

Background: The purpose of this study was to compare a fast spin-echo sequence combined with a respiratory triggering device (R. trig. FSE) with conventional T2-weighted spin-echo (CSE) and inversion recovery (STIR) sequences for the detection of focal hepatic lesions. Methods: We performed a prospective study of 33 consecutive patients with known or suspected hepatic tumors. All patients underwent R. trig. FSE, CSE, and STIR imaging at 1.5 T. Acquisition times were 10.7 min for the CSE sequence and ranged from 12 to 15 min for STIR and from 5 to 7 min for R. trig FSE. For each sequence, liver–spleen contrast-to-noise ratio (CNR) and liver–lesion CNR were determined quantitatively. Image artifact and sharpness were graded by using a four-point scale on each sequence by two independent readers. Both readers also independently identified hepatic lesions (up to a maximum of eight per patient). For patients with focal lesions, the total number of lesions detected (on each sequence) and the minimum size of detected lesions were also determined by each reader. Results: No significant difference was detected between R. trig. FSE and CSE or STIR in either liver–spleen CNR or liver–lesion CNR. R. trig. FSE images were equivalent to CSE and superior to STIR in sharpness (p < 0.01) and presence of artifact (p < 0.01). R. trig. FSE detected a higher number of lesions (reader 1: n = 92, reader 2: n = 86) than CSE (reader 1: n = 70, reader 2: n = 69) and a significantly higher number than STIR (reader 1: n = 71, reader 2: n = 76). Lesion structure was significantly better defined with R. trig. FSE than with STIR (p < 0.01) and CSE (p < 0.05). Conclusions: Compared with CSE and STIR, R. trig. FSE produces hepatic images of comparable resolution and detects an increased number of focal hepatic lesions in a shorter period of time. Received: 5 April 1995/Accepted: 2 May 1995     

Value of endorectal coil versus body coil MRI for diagnosis of recurrent pelvic malignancies

Background: To compare endorectal coil magnetic resonance imaging (MRI) with body coil MRI in detecting local recurrence of gynecologic tumors and prostate and rectal cancers. Methods: Forty-six patients with suspected recurrent pelvic malignancies (13 gynecologic, 15 prostatic, and 18 anorectal primaries) were enrolled in the study. Axial T1- and T2-weighted body coil images and T2- and contrast-enhanced T1-weighted axial endorectal coil images were obtained on a 1.5 T system. Results of the MR examinations were compared with histogical findings and follow-up examinations with respect to the diagnostic accuracy and diagnostic confidence for assessment or exclusion of local recurrence. Results: Recurrent disease was histologically confirmed in eight patients with primary gynecologic malignancies, seven with suspected prostatic recurrence, and seven with suspected anorectal recurrence. Overall, accuracy of body coil MRI was 67% for gynecologic tumors, 36% for prostatic recurrences, and 59% for rectal recurrences. T2- and contrast-enhanced T1-weighted endorectal sequences yielded similar results, with an accuracy of 73% for depiction of gynecologic recurrence, 77% for prostatic recurrence, and 77% for rectal recurrence. The difference in accuracy between body coil and endorectal coil examinations was statistically significant (p < 0.05) only for prostatic cancer. Diagnostic confidence was, however, significantly improved (p < 0.05) in all tumors (T2-weighted endorectal coil examination was superior to T2-weighted body coil images in 71% of cases). Conclusion: Although the results of endorectal coil MRI are only slightly superior to those of body coil MRI for the detection of recurrent gynecologic and anorectal tumors, diagnosis can be made with greater diagnostic confidence in many cases. For detection of prostatic recurrence, endorectal MRI is highly recommended. Received: 27 April 1995/Accepted: 17 June 1995     

Small hepatic nodules in cirrhosis: ultrasonographic, CT, and MR imaging findings

Purpose: The purpose of this study was to assess the imaging findings of pathologically-proved small hepatic nodules 2 cm in size or smaller detected with ultrasonography in cirrhotic patients with suspected hepatocellular carcinoma (HCC). Materials and Methods: We evaluated sonographically detected 32 small hepatic nodules which were pathologically confirmed in 23 consecutive cirrhotic patients who were suspected of having HCC. Twenty-six lesions were confirmed with ultrasonographically-guided aspiration needle-core biopsy, and six with definitive surgery. Ultrasonographic examination records were retrospectively reviewed. CT, and MR images obtained with various imaging techniques were retrospectively reviewed by two radiologists in a blind fashion. Results: The 32 hepatic nodules were comprised of seven focal fatty changes, two large regenerative nodules, three low-grade dysplastic nodules, five high-grade dysplastic nodules, and fifteen HCCs. Ultrasonography showed various echogenicity for the hepatic nodules. The signal-intensity characteristics with T1-weighted spin-echo, in-phase gradient-recalled-echo, and dynamic MR imagings may be useful in distinguishing HCC from nonHCC nodules. Conclusions: Nearly half of small hepatic nodules detected with ultrasonography were nonHCC nodules. Ultrasonographic findings may not be reliable in characterizing small hepatic nodules in cirrhosis. CT and MR imaging obtained with the various techniques are still insensitive to these hepatic nodules. RID="ID="<e5>Correspondence to:</e5> M. Kanematsu Received: 25 August 1997/Revision accepted: 19 November 1997     

Struma ovarii: appearance on MR images

Background: The purpose of this multi-institutional study was to examine the appearance of struma ovarii on magnetic resonance (MR) images. Methods: MR images of 12 patients with histologically proven struma ovarii were retrospectively reviewed. All patients underwent T1-weighted and T2-weighted imaging. Contrast-enhanced T1-weighted images with Gd-DTPA were available in 10 patients. The following determinations were made: tumor morphology, signal intensities, contrast-enhancement effects of solid components with Gd-DTPA, and comparison of MR images with resected specimens. Results: All 12 patients had both cystic and solid components, with a multilobulated surface and thickened septa. Signal intensities on T1-weighted images were mainly low, partly intermediate to high, or high, and those on T2-weighted images were mainly high, with different signal intensities. Contrast-enhancement effects were marked or moderate. The contents that showed low signal intensities on T1-weighted images and signal voids or low signal intensities on T2-weighted images were viscid gelatinous materials. Conclusions: A multicystic tumor with a solid component, a multilobulated surface, and signal intensities that indicate the presence of viscid gelatinous materials appear to be a characteristic MR finding of struma ovarii. Received: 10 April 1997/Accepted after revision: 7 July 1997