| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Developments |
1 From the Department of Radiology, University of Bonn, Sigmund-Freud-Strasse 25, D-53105 Bonn, Germany (N.M., H.H.S., C.K.K., G.L., M.v.F., F.T.); and Philips Medical Systems, Best, the Netherlands (J.G.). Received May 25, 2005; revision requested July 21; revision received December 2; accepted January 6, 2006; final version accepted January 31. Address correspondence to N.M. (e-mail: n.morakkabati{at}uni-bonn.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
© RSNA, 2006
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Single-shot acquisition, the half-Fourier technique, and parallel imaging techniques enable fast imaging acquisition. With these techniques, MR images can be acquired within a short imaging time without artifacts caused by physiologic motion (breathing, motion of the abdominal wall, and peristalsis) and patient movement. With the field strength at 1.5 T, a further reduction in imaging time is limited by an unacceptable degree of signal loss and the resulting impaired image quality.
The currently available higher-field-strength MR imaging systems offer high intrinsic signal-to-noise ratios (SNRs), which potentially enable imaging time to be further shortened. On the other hand, pulse sequence design at high magnetic field strengths faces technical challenges. Especially because of the increased radiofrequency (RF) energy deposition, new strategies are required to reduce the specific absorption rate (1). Furthermore, T1 relaxation time increases while (2,3) T2 relaxation time decreases at 3.0 T (4,5). In addition, insufficient RF power penetration (6), stronger susceptibility effects, and a larger chemical shift have to be taken into account.
The purpose of our study was to prospectively evaluate the diagnostic image quality yielded by a T2-weighted turbo spin-echo sequence with a very short imaging time at 3.0-T MR imaging of the female pelvis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Study Design
From May 2004 through September 2004, we performed a prospective intraindividual study in patients who were referred for MR imaging of the female pelvis. We compared a 3.0-T T2-weighted turbo spin-echo pulse sequence with a very short imaging time and a standard 3.0-T sequence that has been shown to be equivalent to a 1.5-T sequence (7) and which served as a standard of reference. Both T2-weighted turbo spin-echo pulse sequences were performed in a randomized order.
The study design was approved by our institutional review board, and all patients gave informed consent to be examined after the nature of the procedure had been fully explained to them.
Patients
We included 33 consecutive patients (age range, 15–90 years; mean age, 47 ± 16 [standard deviation]). Patients (Table 1) were referred because of a recent diagnosis or history of ovarian tumors (n = 5), a recent diagnosis or history of carcinoma of the cervix (n = 8), a recent diagnosis or history of carcinoma of the vagina (n = 2), uterine malformation (n = 1), indeterminate pelvic mass (n = 2), pelvic pain (n = 4), unknown primary tumor (n = 1), a pelvic lesion at PET/CT (n = 1), or myoma (n = 9). Five patients had undergone a hysterectomy.
|
Intravenous n-butyl-scopolamine (Buscopan; Boehringer-Ingelheim, Ingelheim, Germany) (20 mg) was given to all patients before the examination to reduce peristalsis. In addition, a regional saturation technique device was placed on the anterior abdominal wall to minimize ghosting artifacts caused by the fat in the abdominal wall. MR images of the pelvis were acquired in transverse and sagittal orientations.
The turbo spin-echo pulse sequence we aimed to evaluate is based on a single-shot technique and should have markedly shorter imaging time and identical spatial resolution compared with the standard 3.0-T sequence. The standard sequence was combined with the constant level appearance technique. To reduce the specific absorption rate, the fast sequence was combined with the half-Fourier technique, parallel imaging (sensitivity encoding [SENSE]), and a variable refocusing angle technique (8–12) called the flip-angle sweep technique that uses RF pulses with lower power deposition.
The imaging time for the fast sequence was reduced to 39 seconds, as compared with 4 minutes 3 seconds for the standard sequence.
The following parameters were increased with the fast sequence (Table 2): Two signals were acquired (vs one with the standard sequence), the turbo factor was 73 (vs 25 with the standard sequence), the repetition time was 4933 msec (vs 2705 msec with the standard sequence), and the echo time was 100 msec (vs 80 msec with the standard sequence).
|
We qualitatively evaluated tissue contrast for anatomic structures such as the zonal anatomy of the uterus (the sharp border between the hypointense junctional zone and the intermediate signal intensity of myometrium)—except in patients with hysterectomy or tumor invasion into the uterus—the ovarian stroma and ovarian cysts, and muscle and urine. In addition, in the setting of gynecologic disorders, we evaluated contrast between muscle and solid tumors and contrast between muscle and cystic tumors or liquids (eg, ascites, hydronephrosis).
Again, a three-point scale was used. We evaluated whether contrast was higher (3 points), equal (2 points), or lower (1 point) on images obtained with the fast sequence than on images obtained with the standard sequence. Concerning the zonal uterine anatomy, 3 points were assigned if the junctional zone appeared somewhat more hypointense or if the myometrium appeared more hyperintense on images obtained with the fast sequence. Two points were assigned if the signal intensities of the hypointense junctional zone and the intermediate myometrium were comparable. One point was assigned if the junctional zone appeared somewhat less hypointense or if the myometrium appeared less hyperintense on images obtained with the fast sequence.
In addition, ROI-based quantitative measurements were performed (N.M.) to assess different tissue contrasts (C) according to the following equation: C = (A – B)/(A + B), where A represents the signal intensity of tissue A and B represents the signal intensity of tissue B.
For quantitative evaluation, the largest possible ROI was placed. It was not possible to copy the ROI from the corresponding MR images because the uterus and the ovaries are not static organs but move to some extent. Moreover, bladder filling during the course of the examination may also alter the location of the uterus and the ovaries. To ensure consistency, ROIs were carefully placed by the same author in the corresponding anatomic regions (this was done by visual comparison—eg, by considering neighboring anatomic landmarks or the shape of a tumor).
The size of the ROIs varied depending on the anatomic location. Minimum size was 7 mm2 for ovarian stroma and cysts, and maximum size was 100 mm2 in tumors.
ROIs were placed in gluteal muscle, carcinomas of the cervix, uterine myomas, solid and cystic parts of ovarian tumors, ovarian stroma, ovarian cysts, the junctional zone, the myometrium, and the bladder.
We analyzed contrast between muscle and solid tumors, contrast between the junctional zone and the uterine myometrium (zonal uterine anatomy), contrast between muscle and cystic tumors or liquids (eg, ascites and hydronephrosis), contrast between muscle and urine, and contrast between ovarian cysts and ovarian stroma.
The degree of artifacts due to ghosting of the abdominal wall and peristalsis was evaluated by two radiologists (N.M. and M.v.F.) in consensus by using a five-point scale. The images obtained with the fast sequence and the corresponding images obtained with the standard sequence were evaluated separately.
One point was assigned if no artifacts were present. Two points were assigned if minor artifacts were present—that is, if only some small bright or dark repeating lines oriented in the phase-encoding direction were visible or if the border of the bowel appeared slightly indistinct.
Three points were assigned if moderate (not diagnostically relevant) artifacts were present—that is, if the repeating lines appeared somewhat more distinct or if the border of the bowel appeared somewhat more indistinct but the anatomic structures could still be identified well and the interpretation of the images was not markedly affected.
Four points were assigned if stronger (diagnostically relevant) artifacts were present—that is, if anatomic structures and gynecologic disorders were difficult to identify and interpretation of the images was markedly affected.
Five points were assigned if severe artifacts were present (nondiagnostic study)—that is, the images could not be interpreted.
Delineation and detectability of image details were evaluated with respect to the visualization of small anatomic details (fine septa in the fat tissue, small vessels [if present], small ovarian cysts, and [if possible] the zonal anatomy of the uterus). Furthermore, the detectability of gynecologic disorders was evaluated. Two radiologists (N.M. and M.v.F.) compared the MR images acquired with the fast and standard sequences directly in consensus by using a three-point scale.
Three points were assigned if the borders between fat septa and fat tissue, between ovarian cysts and ovarian stroma, between small vessel walls and fat tissue, and between gynecologic tumors and neighboring structures appeared somewhat sharper on images obtained with the fast sequence or if more details (eg, small cysts, fat septa) were visible on these images. Two points were assigned if no differences could be detected between images acquired with the fast sequence and those acquired with the standard sequence. One point was assigned if borders appeared somewhat sharper or if more details were visible on images acquired with the standard sequence.
We compared the final MR imaging diagnoses that were obtained with the standard pulse sequence and the fast pulse sequence. Images obtained with each sequence were evaluated separately, with the readers blinded to the sequence used. We applied the standard MR imaging criteria that are in use for gynecologic disorders of the female pelvis: For myomas, we evaluated their size, number, and location. In case of uterine malformation, we used the American Fertility Society classification of müllerian duct anomalies. The staging of carcinoma of the cervix and carcinoma of the ovary was performed with the Fédération Internationale de Gynécologie et d'Obstétrique classification.
Statistical Analysis
For statistical analysis, a software package (SPSS; SPSS, Chicago, Ill) was used to calculate mean values and standard deviations for tissue contrast measurements. To test for statistical significance, we used the Wilcoxon paired test for quantitative analysis of tissue contrasts. We used the McNemar test for the qualitative analysis of tissue contrasts. The marginal homogeneity test was used to test for statistically significant differences in artifact levels (according to scores on the five-point scale) between the fast and standard sequences. P < .05 was considered to indicate a significant difference for all analyses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
In 11 of 33 patients (33%), we could not analyze the tissue contrast of the zonal anatomy. Five of these 11 patients had undergone hysterectomy, and in six of these 11 patients, the zonal anatomy had been destroyed by tumor invasion. For the remaining 22 patients, qualitative analysis revealed comparable tissue contrast (mean score, 2 points) for the zonal uterine anatomy with the standard and fast sequences.
We diagnosed 27 solid tumors in 24 patients and 12 cystic tumors or liquid-containing disorders in 10 patients (Table 3).
|
|
|
|
|
|
|
|
With the standard sequence, artifacts were moderate in 10 of 33 patients, minimal in 17 of 33 patients, and absent in six of 33 patients.
Chemical shift artifacts or dielectric artifacts were not observed in this study.
The delineation and detectability of small anatomic details and of gynecologic disorders was rated equal for all (33 of 33) pelvic MR imaging studies; final MR imaging diagnoses did not differ between the two sequences.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Fast MR imaging allows increased patient throughput. Furthermore, owing to the reduced imaging time, fast MR sequences allow minimization of motion artifacts, which is especially important in uncooperative patients. In our study, we compared a 3.0-T T2-weighted turbo spin-echo sequence with a very short imaging time with a standard 3.0-T T2-weighted turbo spin-echo pulse sequence in female patients undergoing MR imaging of the pelvis because of various gynecologic disorders.
It has to be stated that, in principle, the higher signal at 3.0 T enables a further decrease in imaging time. Because of the altered specific absorption rate conditions at 3.0 T, however, several technical modifications have to be considered; this is even more important with a single-shot turbo spin-echo sequence. Because specific absorption rate limits are reached earlier at 3.0 T, the echo train length of a single-shot sequence—which leads to a very high energy deposition due to the high number of 180° RF pulses—will need to be limited to avoid heating and blurring. Because we wanted to maintain the spatial resolution of the standard pulse sequence, strategies to markedly reduce the energy deposition were needed. The new pulse sequence was first combined with the half-Fourier technique and with a parallel imaging technique (ie, SENSE), which shorten the echo train length and thus reduce energy deposition and imaging time. Furthermore, SENSE was needed to minimize possible blurring artifacts, which may occur with a single-shot technique because of high turbo factors. This issue is even more important at 3.0 T because the high field strength itself is more susceptible to blurring artifacts.
With use of a SENSE factor of three, energy deposition was still too high and imaging time was still too long. We did not use a higher SENSE factor because in this situation signal would have decreased further and infolding artifacts might have occurred, thus impairing image quality.
Instead, the fast sequence was combined with a variable refocusing angle technique (8–12), which uses RF pulses with lower power deposition. The variable refocusing angle technique produces a higher amount of stimulated echoes as compared with techniques involving 180° refocusing pulses only. It has been described (8,9) that the variable refocusing angle technique thus results in a reduction of effective echo time, meaning that a longer echo time will be necessary to reach adequate T2 weighting. The choice of the longer echo time for the fast sequence was based on prior measurements.
To maintain the familiar tissue contrasts of a standard sequence, repetition time also has to be adapted when the variable refocusing angle technique is used. In detail, this technique requires a repetition time that is long enough to collect a considerable amount of stimulated echoes, thus increasing signal and pronouncing T2 weighting. In contrast to Hennig et al (8,9), who used the hyperecho technique, the flip-angle sweep technique can be combined with a single-shot pulse sequence.
In summary, the high field strength at 3.0 T enabled the substantial reduction of imaging time with a T2-weighted turbo spin-echo pulse sequence for MR imaging of the female pelvis.
To assess whether this pulse sequence maintained the diagnostic image quality of a standard sequence, we analyzed the visual SNRs and tissue contrasts of anatomic structures and frequent gynecologic disorders. In addition, we wondered if the fast MR imaging technique significantly reduces motion artifacts. Regarding the apparent SNR, we did not encounter significant differences between the two sequences, although we had applied several techniques that could have led to a visible signal loss with the fast sequence. Obviously, the choice of a longer repetition time and the increased number of signals acquired was able to compensate for a potential signal loss caused by SENSE, the half-Fourier technique, the flip-angle sweep technique, and the longer echo time. We did not reduce the SENSE factor—which would have reduced signal less—but preferred to increase the number of signals acquired. This way of proceeding may appear contradictory at the first glance. But after a closer look, it becomes obvious that the use of SENSE was mandatory—first, to reduce blurring with a single-shot technique, and second, to reduce energy deposition drastically. Regarding whether the fast sequence also provided familiar tissue contrasts, our data show that the majority of tissue contrasts were comparable for both sequences.
With respect to liquids, tissue contrast was even higher with the fast sequence. This may be explained by the use of the flip-angle sweep technique. Hennig et al (8,9) report that for liquids (with T1 = T2), the effective echo time will remain constant and will not be affected by the amount of stimulated echo contributions. Therefore, the choice of a longer echo time will increase T2 weighting for tissues containing liquids. This fact, however, did not disturb image interpretation. On the contrary, in our small study cohort, the detection of fluid collections was improved with the new sequence owing to the higher signal. Moreover, with the chosen contrast parameters, we were also able to detect and diagnose all gynecologic pelvic disorders with the fast sequence.
Concerning the amount of motion artifacts, our data confirm that the fast sequence was superior to the standard sequence. Theoretically, the risk of patient movement is increased sixfold with the standard protocol owing to the sixfold longer imaging time. Although we did not observe severe patient movement in our study cohort, it is conceivable that the fast sequence will be advantageous in uncooperative patients. Physiologic motion artifacts due to breathing—and motion of the abdominal wall—or peristalsis were drastically reduced with the fast MR imaging sequence. Further artifacts like blurring or chemical shift were not observed in this study. Obviously, parallel imaging was able to prevent an increase in blurring with a single-shot technique at high field strength.
To evaluate the diagnostic potential of the fast sequence, we analyzed delineation and detectability of small anatomic details and of gynecologic disorders. All small anatomic details were detected with both sequences. With regard to gynecologic disorders, all solid (n = 27) and cystic (n = 12) pelvic disorders that had been diagnosed with the standard sequence could also be diagnosed with the fast sequence. In some cases, delineation of tumor margins was even better with the fast sequence owing to a reduction in motion artifacts.
In summary, the fast sequence provided comparable visual SNRs, significantly reduced artifacts, and maintained or increased tissue contrasts, thus maintaining or improving the diagnostic image quality yielded by a standard sequence.
Our study was limited because data analysis was performed in consensus. Thus, we are not able to provide data for interobserver variability.
In conclusion, a flip-angle sweep technique and SENSE enable fast MR imaging at 3.0 T while maintaining or improving the diagnostic image quality provided by a standard pulse sequence. To the best of our knowledge, ours is the first report on this technique for MR imaging of the female pelvis. The fast sequence we describe may be advantageous for pelvic MR imaging of uncooperative patients and patients with contraindications to intravenous n-butyl-scopolamine. It can enable an increase in patient throughput and seems to be suited for the acquisition of additional anatomic views.
Therefore, we recommend use of the fast sequence routinely instead of the standard sequence.
|
ADVANCES IN KNOWLEDGE |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
|
FOOTNOTES |
|---|
Abbreviations: RF = radiofrequency • ROI = region of interest • SENSE = sensitivity encoding • SNR = signal-to-noise ratio
See Materials and Methods for pertinent disclosures.
Author contributions: Guarantors of integrity of entire study, N.M., H.H.S., J.G.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, N.M.; clinical studies, N.M., C.K.K., G.L., M.v.F., J.G.; statistical analysis, F.T.; and manuscript editing, all authors
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. Hottat, C. Larrousse, V. Anaf, J.-C. Noel, C. Matos, J. Absil, and T. Metens Endometriosis: Contribution of 3.0-T Pelvic MR Imaging in Preoperative Assessment--Initial Results Radiology, July 7, 2009; (2009) 2531082113. [Abstract] [Full Text]
|
|
|
|
|
|
M. Hori, T. Kim, T. Murakami, I. Imaoka, H. Onishi, K. Tomoda, T. Tsutsui, T. Enomoto, T. Kimura, and H. Nakamura Uterine Cervical Carcinoma: Preoperative Staging with 3.0-T MR Imaging--Comparison with 1.5-T MR Imaging Radiology, April 1, 2009; 251(1): 96 - 104. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. K. Kuhl, F. Traber, J. Gieseke, W. Drahanowsky, N. Morakkabati-Spitz, W. Willinek, M. von Falkenhausen, C. Manka, and H. H. Schild Whole-Body High-Field-Strength (3.0-T) MR Imaging in Clinical Practice * Part II. Technical Considerations and Clinical Applications Radiology, April 1, 2008; 247(1): 16 - 35. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. K. Kuhl, F. Traber, and H. H. Schild Whole-Body High-Field-Strength (3.0-T) MR Imaging in Clinical Practice * Part I. Technical Considerations and Clinical Applications Radiology, March 1, 2008; 246(3): 675 - 696. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
