AJR ARRS: Your Link to CME
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Medved, M.
Right arrow Articles by Karczmar, G. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Medved, M.
Right arrow Articles by Karczmar, G. S.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
DOI:10.2214/AJR.04.1704
AJR 2006; 186:30-37
© American Roentgen Ray Society

Original Research

High Spectral and Spatial Resolution MRI of Breast Lesions: Preliminary Clinical Experience

Milica Medved1, Gillian M. Newstead1, Hiroyuki Abe1, Marta A. Zamora1, Olufunmilayo I. Olopade2 and Gregory S. Karczmar1

1 Department of Radiology, The University of Chicago, 5841 S Maryland Ave., MC 2026, Chicago, IL 60637.
2 Section of Hematology and Oncology, The University of Chicago, Chicago, IL 60637.

Received November 2, 2004; accepted after revision January 4, 2005.

 
Address correspondence to G. S. Karczmar (gskarczm{at}uchicago.edu).

Supported by grants from the National Institute of Biomedical Imaging and Bioengineering (RO1 EB003108-01), the National Cancer Institute (RO1 CA78803), the Army Breast Cancer Research Program (DAMD 17-02-1-0033), GE Healthcare, the Segal Foundation, the Falk Medical Research Trust, and the Doris Duke Charitable Foundation.


Abstract
Top
 Abstract
Introduction
 Subjects and Methods
 Results
 Discussion
 References
 
OBJECTIVE. In previous research, high spectral and spatial resolution (HiSS) echo-planar spectroscopic imaging (EPSI) was successfully applied to the human breast, obtaining improved contrast, anatomic detail, and sensitivity to contrast agents. To test HiSS in the clinical setting, we used HiSS MRI to image 30 women with suspicious breast lesions.

SUBJECTS AND METHODS. Women with suspicious breast lesions were scanned before and after contrast administration using EPSI at 1.5 T (0.63-mm in-plane resolution, 2.6-Hz spectral resolution). Images with intensity proportional to the water signal peak height in each voxel were synthesized and compared with standard clinical fat-saturated and early dynamic subtraction images. Pre- and postcontrast HiSS images were compared to assess the effect of the contrast agent on water resonance structure.

RESULTS. HiSS images scored significantly better than standard clinical images in lesion conspicuity, margin definition, and internal definition, even though they were acquired before contrast agent injection. Fat suppression was more complete and uniform and detail was shown on HiSS images more clearly than on conventional fat-saturation images. Thus, HiSS images often allowed easier evaluation of the lesion. Contrast agent-affected changes were often spatially and spectrally inhomogeneous.

CONCLUSION. HiSS scans were successfully integrated into standard clinical examinations and provided diagnostically useful images before contrast agent injection. Thus, it might be possible to characterize suspicious lesions on the basis of precontrast high-resolution spectral information. This information and information about the effect of contrast agents could potentially improve the specificity of breast MRI.

Keywords: breast • breast cancer • high spectral and spatial resolution spectroscopic imaging • MRI • screening


Introduction
Top
 Abstract
 Introduction
Subjects and Methods
 Results
 Discussion
 References
 
Over the past decade, breast MRI has become established as a useful tool in both screening and diagnosis of breast cancer [1-3]. Two criteria found to be the most important for differentiation of malignant from benign lesions are their morphology and rate of contrast uptake [4-7]. However, these two criteria provide limited accuracy, mainly because of the conflicting requirements of high spatial resolution and high temporal resolution on the imaging sequence. These demands may prove difficult to reconcile, and obtaining definitive clinical information during a single examination may require new types of MR contrast to be developed for breast MRI. Several techniques are being explored, including diffusion and perfusion imaging [8, 9], elastography [10], equivalent cross-relaxation rate imaging [11], and postprocessing techniques [12-15]. Although these methods are promising, none has yet shown adequate sensitivity and specificity for detection of early cancers. Therefore, high spectral and spatial resolution (HiSS) MRI was developed to improve functional and anatomic measurements and enhance the clinical utility of MRI for the detection and characterization of breast cancer [16-19].

HiSS imaging is currently acquired at our laboratory with a high-resolution echo-planar spectroscopic imaging (ESPI) sequence [20]. Quantitative comparison of HiSS MR images and conventional images suggests that in the breast, HiSS imaging improves fat suppression, contrast, and anatomic detail [16, 17]. Its increased sensitivity to contrast agents [17] and the ability to resolve the fine structure of the water line [16, 18] indicate that HiSS imaging may provide a new type of contrast based, for example, on imaging individual Fourier components of the water resonance that correspond to different subvoxel water compartments in the tissue [18].

An important step in developing HiSS imaging as a source of new contrast in breast MRI is its successful incorporation into the standard clinical scans. To date, there have been few images of small breast lesions that provide a realistic test of HiSS in the clinical setting, where challenges include proper slice positioning, maintaining the clinical schedule, technologist training, and so on. Here, we report on the use of HiSS MRI to image 30 women with suspicious breast lesions. We evaluate the practicality of implementing HiSS imaging in the clinical setting and compare fat-suppressed precontrast HiSS images with clinical fat-saturated postcontrast images and early dynamic subtraction images to assess their quality and diagnostic utility. In addition, we report on the effects of contrast agents on the shape of the water resonance line in small voxels.


Subjects and Methods
Top
 Abstract
 Introduction
 Subjects and Methods
Results
 Discussion
 References
 
Patient Recruitment
Thirty women with suspicious breast lesions detected on X-ray mammography or physical examination were scanned. To obtain a more uniform patient population, we excluded lesions with associated microcalcifications. The biopsy-determined final diagnoses for the imaged patients are summarized in Table 1. The patients were recruited from the University Cancer Risk Clinic and the Breast Imaging Service and were studied under a protocol approved by the university institutional review board after informed consent was obtained from study participants.


View this table:
[in this window]
[in a new window]

  		TABLE 1: Biopsy-Confirmed Diagnoses in Patient Population

 

Image Acquisition
MR images were acquired on 1.5-T scanners (Signa, GE Healthcare) equipped with EchoSpeed gradients (GE Healthcare) with a maximum slew rate of 120 mT/m/sec and maximum amplitude of 23 mT/m. The body coil was used for excitation, and a phased-array coil (4-element Breast Array, GE Healthcare) was used to detect signal. Light compression of the breasts was applied to reduce motion artifacts. HiSS images were acquired using EPSI [17, 20]. The EPSI pulse sequence was composed of a slice-selective excitation pulse (4-mm slice thickness), a phase-encoding gradient with 256 phase-encoding steps, and a train of 128 gradient echoes with 384 sampled points per gradient echo. This produced images with an in-plane resolution of 0.63 mm and 128 bins of spectral resolution. The gradient-echo train was sampled at a bandwidth of 62.5 kHz, so that the free induction decay (FID) was sampled for 384 msec, and the spacing between echoes was 3.0 msec. This resulted in a spectral resolution of 2.6 Hz, or 0.041 ppm, and a spectral bandwidth of 333 Hz. The acquisition time per slice under these conditions was 128 sec. Typically, only a single slice was imaged in each patient. Because the lesion was not always visible in the axial T2-weighted sequence, previous mammographic or MRI examinations (or both) were used to position the HiSS slice. Shimming was performed on the slice immediately before HiSS imaging. The patients were imaged both before and after contrast agent injection (either 0.1 mmol/L per kilogram of patient body weight or 20 mL of gadodiamide [Omniscan, Nycomed Amersham] injected at rate of 2 mL/sec). The shim and gain parameters were identical for both HiSS scans.

Imaging Protocol
The imaging protocol included an axial fast spin-echo acquisition using the following parameters: TR/TE, 8,500/102 msec; bandwidth, 41.67 MHz; slice thickness, 5 mm; interleaved slice acquisition; matrix size, 256 x 224; and number of excitations (NEX), 2. An unenhanced sagittal HiSS acquisition was then performed. (All HiSS acquisitions were performed with parameters described in the "Image Acquisition" section of this article.) An unenhanced 3D coronal fast spoiled gradient-echo (SPGR) acquisition with a TE of 4.7 msec to obtain fat and water signals in-phase (flip angle, 30°; bandwidth, 31.25 MHz; slice thickness, 3 mm; matrix size, 256 x 224; NEX, 1) was performed (duration, 68 sec). All subsequent 3D coronal fast SPGR acquisitions were obtained with the same parameters. Next, contrast agent was injected and followed by a 20-sec delay. A postcontrast 3D coronal fast SPGR acquisition was repeated twice, followed by a postcontrast sagittal HiSS acquisition and a postcontrast 3D coronal fast SPGR acquisition. A postcontrast bilateral axial fast SPGR acquisition was then performed (TR/TE, 150/4.7; flip angle, 80°; bandwidth, 20.83 MHz; matrix size, 256 x 224; NEX, 2; and slice thickness, 5 mm). Finally, a unilateral fat-saturated sagittal fast SPGR acquisition on one or both breasts was performed using the following parameters: TR/TE, 175/4.7; flip angle, 80°; bandwidth, 31.25 MHz.



View larger version (14K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1 —Histogram shows sizes of lesions for 30 imaged patients. Lesion size ranged from 7 to 58 mm, and average lesion size was 22 mm.

 



View larger version (77K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2A —Patient 1: 67-year-old woman with noncalcified fibrocystic change with usual ductal hyperplasia nodule. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 



View larger version (99K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2B —Patient 1: 67-year-old woman with noncalcified fibrocystic change with usual ductal hyperplasia nodule. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 



View larger version (76K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2C —Patient 1: 67-year-old woman with noncalcified fibrocystic change with usual ductal hyperplasia nodule. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 



View larger version (93K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3A —Patient 2: 41-year-old woman presenting with mammographically detected grade II ductal carcinoma with ductal carcinoma in situ. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 



View larger version (131K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3B —Patient 2: 41-year-old woman presenting with mammographically detected grade II ductal carcinoma with ductal carcinoma in situ. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 



View larger version (104K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3C —Patient 2: 41-year-old woman presenting with mammographically detected grade II ductal carcinoma with ductal carcinoma in situ. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 
Image Processing
HiSS data were processed by applying a 3D Fourier transform to the raw data to obtain a high-resolution proton spectrum in every voxel of the imaged slice. An algorithm described in detail in an earlier article [16] was used on the resulting spectra to detect the frequencies of the fat and water peaks. The high-resolution magnitude spectrum in each voxel was fitted using a Lorentzian function approximation for the fat and water peaks and for any N/2 (Nyquist) spectral ghost peaks. The Lorentzian fits for fat peaks and N/2 ghost peaks and any constant background signal were subtracted to reveal the pure water signal. This was particularly important for the detection of small amounts of water in the presence of large fat signals. From the resulting water spectra, water peak height images were constructed. Because the water signal was often inhomogeneously broadened, the peak height images were not obtained from the fit parameters. Rather, the maximum value of the isolated water signal in each voxel was used to construct water peak height images. All processing software was developed in Interactive Data Language (IDL, Research Systems). The images presented were processed in approximately 45 min per image on a 3.2-GHz PC (Pentium 4 CPU, Intel).

The kinetic data were obtained by manually selecting an oval region of interest (ROI) within the lesion and categorized by a radiologist. The sagittal dynamic contrast-enhanced MR subtraction images shown here were reconstructed from the coronal 3D acquisition sequence. The ROI selection (before reconstruction) and the reconstruction were performed on a workstation (Advantage 4.0, GE Healthcare).

Image Scoring
Two experienced radiologists rated the quality of HiSS images relative to standard clinical MR images for lesion conspicuity and for internal and margin definition. The scale used was –2, much worse than standard MR images; –1, worse than; 0, equal to; 1, better than; and 2, much better than standard MR images. The consensus opinion was reported. Before comparison, HiSS images were down-sampled, using cubic interpolation, to the same nominal spatial resolution as standard clinical images to eliminate any bias.



View larger version (155K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4A —Patient 3: 67-year-old woman with invasive lobular carcinoma. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 



View larger version (143K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4B —Patient 3: 67-year-old woman with invasive lobular carcinoma. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 



View larger version (88K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4C —Patient 3: 67-year-old woman with invasive lobular carcinoma. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 



View larger version (103K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5A —Patient 4: 66-year-old woman with infiltrating ductal carcinoma mass lesion. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 



View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5B —Patient 4: 66-year-old woman with infiltrating ductal carcinoma mass lesion. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 



View larger version (85K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 5C —Patient 4: 66-year-old woman with infiltrating ductal carcinoma mass lesion. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 

Results
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
Discussion
 References
 
Figure 1 shows the distribution of all lesion sizes, defined as the longest linear dimension when measured in three planes on standard clinical images. The lesions ranged from 7 to 58 mm, and the average size was 22 mm. For two of the patients, there was only parenchymal enhancement (no lesion) on MRI examination, and in one patient, the HiSS slice was not properly positioned to include the lesion. In three other patients, the postcontrast T1-weighted fat-saturated images were not acquired or were of poor quality because of technical problems or patient motion or refusal. All lesions visible on clinical examination were also visible on corresponding HiSS images. Thus, images for 24 patients were available for comparison of HiSS imaging with standard clinical imaging.

Figures 2A, 2B, 2C, 3A, 3B, 3C, 4A, 4B, 4C, 5A, 5B, 5C, 6A, 6B, and 6C show the standard clinical postcontrast T1-weighted fat-saturated images (Figs. 2A, 3A, 4A, 5A, and 6A), HiSS images (Figs. 2B, 3B, 4B, 5B, and 6B), and the dynamic contrast-enhanced MRI subtraction images at the second minute postinjection (Figs. 2C, 3C, 4C, 5C, and 6C) for six patients. The HiSS images are displayed at their true in-plane resolution of 0.63 mm. In all patients, fat suppression was superior and more uniform and image contrast is improved on HiSS images. Next, we summarize the cases shown in Figures 2A, 2B, 2C, 3A, 3B, 3C, 4A, 4B, 4C, 5A, 5B, 5C, 6A, 6B, and 6C.



View larger version (64K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 6A —Patient 5: 48-year-old woman with ductal carcinoma in situ. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 


View larger version (108K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 6B —Patient 5: 48-year-old woman with ductal carcinoma in situ. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 


View larger version (92K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 6C —Patient 5: 48-year-old woman with ductal carcinoma in situ. Standard clinical postcontrast T1-weighted fat-saturated image (A), high spectral and spatial resolution (HiSS) image (B), and dynamic contrast-enhanced MR subtraction image at second minute postinjection (C) in sagittal projection are displayed.

 
Patient 1 was an asymptomatic 67-year-old woman who presented for annual mammography. The mammograms showed a 9-mm ill-defined noncalcified nodule at the 6-o'clock position in the right breast with sonographic correlation. Conventional MRI depicted a 7-mm homogeneously enhancing nodule at the 6-o'clock position of the right breast. Kinetic data for this nodule showed a medium rise with washout pattern. HiSS images better depicted margin definition and internal definition of this lesion than the postcontrast fat-saturated T1-weighted images. Sonographically guided core needle biopsy revealed fibrocystic change with usual ductal hyperplasia.

Patient 2 was a 41-year-old woman who presented with a mammographically detected 10-mm spiculated mass in the left breast. Conventional MRI showed an irregular enhancing mass in the left upper outer quadrant, 10 mm in size. Kinetic data for the mass showed rapid rise with washout. HiSS images depicted the lesion with higher conspicuity and with better defined margins on both pre- and postcontrast image sets than the postcontrast fat-saturated T1-weighted images. Excisional biopsy indicated a 1.4-cm grade II invasive ductal carcinoma with ductal carcinoma in situ.

Patient 3 was a 67-year-old woman who presented with a 1.2-cm spiculated mass at the 9-o'clock position in the right breast seen on mammography and sonography examinations. Conventional MRI revealed an irregular mass with heterogeneous internal enhancement and spiculations visible on postcontrast fat-saturated T1-weighted images. Kinetic data for the mass showed rapid washout. HiSS data revealed superior margin visualization, with clearly evident spiculations. Biopsy showed a 1.2-cm invasive lobular carcinoma.

Patient 4 was an asymptomatic 66-year-old woman who presented with a new 1-cm mass lesion in the upper inner quadrant of the left breast detected on an annual screening mammogram. The mass was ill-defined and noncalcified and was correlated with a sonographic nodule with minimal shadowing. Conventional MRI showed a 1-cm enhancing lesion with the kinetic pattern of rapid uptake and slow washout. HiSS images depicted margin definition and internal definition of this lesion better than postcontrast fat-saturated T1-weighted images. Sonographically guided core needle biopsy revealed infiltrating ductal carcinoma.

Patient 5 was a 48-year-old woman with a history of intermittent clear and bloody nipple discharge with a palpable abnormality in the right breast. The mammogram failed to show an abnormality, but sonography showed an ill-defined hypoechoic lesion that was 1.3 cm with internal blood flow. Conventional MRI revealed a 40 x 30 mm area of irregular enhancement at the site of palpable abnormality, with the kinetic pattern of fast uptake and washout. HiSS images depicted a linear lesion ({approx} 30 mm in plane) with multiple foci at the area of abnormal enhancement on the conventional MR images. HiSS images had improved conspicuity and depicted the margin definition and internal definition of this lesion better than postcontrast fat-saturated T1-weighted images. Surgical excision of this area was performed, and pathology revealed ductal carcinoma in situ measuring at least 15 mm.

Table 2 summarizes the radiologists' scoring (average scores ± SD) relative to the standard images for all lesions, malignant invasive lesions, ductal carcinomas in situ (DCIS), and benign lesions with the number of patients in each group. On average, HiSS images scored significantly better than standard clinical images in all categories based on a one-sided Student's t test (p < 0.01). For DCIS and benign lesions, the sample size was small, so this analysis was not performed. The percentages of patients for whom HiSS images were assigned higher scores than clinical images were 63%, 79%, and 96% for lesion conspicuity, margin definition, and internal definition, respectively. The percentages of patients for whom HiSS images were assigned scores equal to or better than clinical images were 79%, 88%, and 96%, respectively. Only the results for invasive malignant lesions were similar.


View this table:
[in this window]
[in a new window]

  		TABLE 2: High Spectral and Spatial Resolution (HiSS) MR Image Scores Relative to Standard Clinical Images, per Diagnostic Category

 

The effect of contrast agents on HiSS water peak height images is frequently spatially and spectrally inhomogeneous and is therefore a potential source of novel MRI contrast. This is illustrated by spectral data from patient 6—a 40-year-old woman at high-risk for breast cancer who presented with a palpable mass at the 12-o'clock position in the posterior left breast. Her mammography and sonography examinations showed a 2.5-cm irregular mass consistent with malignancy. Palpable nodes were noted clinically in the left axilla. Conventional MR images showed a 28-mm irregular heterogeneously enhancing index mass with a second anterior 10-mm focus of tumor. Kinetic data for the index lesion showed fast uptake with washout. HiSS images depicted an irregular mass with improved definition of the internal enhancement characteristics. Core biopsy of the index lesion revealed a high-grade invasive ductal carcinoma.

Figures 7A, 7B, 7C, 7D, 7E, and 7F shows the conventional postcontrast T1-weighted MR image (Fig. 7A), precontrast HiSS images displayed using different window settings (Figs. 7B and 7C), and the difference between 3-min post- and precontrast HiSS images (Fig. 7D) for patient 6. In Figure 7B, the window settings were chosen to show general breast anatomy, whereas the window settings were optimized to show inherent contrast within the lesion in Figure 7C. The two different window settings are necessary because of the high dynamic range of HiSS water peak height images. Figure 7D—the image depicting the difference between the pre- and postcontrast HiSS images—shows spatial inhomogeneity of the contrast agent effect on water resonance, observed at 3 min after injection, and resulting in peak height decreases (dark) in some and increases (bright) in other regions. The spectral inhomogeneity in the contrast agent effect is illustrated in Figures 7E and 7F, where the water resonance measured before (dashed line in Figs. 7E and 7F) and after (solid line in Figs. 7E and 7F) contrast injection is compared for two voxels indicated by arrows in Figure 7D. In the voxel depicted in Figure 7E, there appears to be a single water resonance that is slightly shifted and homogeneously broadened after contrast agent administration. In the voxel depicted in Figure 7F, two components can be observed; one is broadened and shifted, and the other shows a small increase in height after contrast agent administration. The changes observed are above the noise level and can be used as a source of MR contrast.



View larger version (108K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 7A —Patient 6: 40-year-old woman at high risk for breast cancer who presented with palpable invasive ductal carcinoma mass. All images are shown in sagittal projection. Conventional postcontrast T1-weighted MR image.

 


View larger version (158K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 7B —Patient 6: 40-year-old woman at high risk for breast cancer who presented with palpable invasive ductal carcinoma mass. All images are shown in sagittal projection. Precontrast high spectral and spatial resolution (HiSS) images of same slice as A displayed using different window settings to show general breast anatomy (B) and to show inherent contrast within lesion (C).

 


View larger version (93K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 7C —Patient 6: 40-year-old woman at high risk for breast cancer who presented with palpable invasive ductal carcinoma mass. All images are shown in sagittal projection. Precontrast high spectral and spatial resolution (HiSS) images of same slice as A displayed using different window settings to show general breast anatomy (B) and to show inherent contrast within lesion (C).

 


View larger version (117K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 7D —Patient 6: 40-year-old woman at high risk for breast cancer who presented with palpable invasive ductal carcinoma mass. All images are shown in sagittal projection. Image shows difference between 3-min postcontrast and precontrast HiSS images. Image reveals spatial inhomogeneity of contrast agent effect on water resonance that was observed at 3 min after injection and resulted in peak height decreases (dark) in some and increases (bright) in other regions. Arrows point to 2 voxels for which water resonance is shown in E and F.

 


View larger version (13K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 7E —Patient 6: 40-year-old woman at high risk for breast cancer who presented with palpable invasive ductal carcinoma mass. All images are shown in sagittal projection. Images illustrate water resonance measured for 2 voxels indicated by arrows in D before (dashed line) and after (solid line) administration of contrast material for comparison. In the voxel depicted in E, there appears to be a single water resonance that is slightly shifted and homogeneously broadened after contrast agent administration. In the voxel depicted in F, two components can be observed; one is broadened and shifted, and the other shows a small increase in height after contrast agent administration. The changes observed are above the noise level and can be used as a source of MR contrast. E and F are shown on an arbitrary scale.

 


View larger version (13K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 7F —Patient 6: 40-year-old woman at high risk for breast cancer who presented with palpable invasive ductal carcinoma mass. All images are shown in sagittal projection. Images illustrate water resonance measured for 2 voxels indicated by arrows in D before (dashed line) and after (solid line) administration of contrast material for comparison. In the voxel depicted in E, there appears to be a single water resonance that is slightly shifted and homogeneously broadened after contrast agent administration. In the voxel depicted in F, two components can be observed; one is broadened and shifted, and the other shows a small increase in height after contrast agent administration. The changes observed are above the noise level and can be used as a source of MR contrast. E and F are shown on an arbitrary scale.

 

Discussion
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
References
 
The data demonstrate the feasibility of imaging small (down to 7 mm) breast lesions in a standard clinical setting using the HiSS sequence, even when only a single 4-mm-thick slice is acquired. HiSS images scored significantly better than the standard clinical images on lesion conspicuity, margin definition, and internal lesion definition. This often allowed easier lesion evaluation, even though precontrast HiSS images were compared with postcontrast conventional images. The results indicate that HiSS MRI could be used to detect and possibly to characterize lesions before contrast administration on the basis of margin analysis and internal definition.

The equal or higher quality of precontrast HiSS images compared with postcontrast conventional images is not due to their higher spatial resolution, because radiologists evaluated water peak height images that were down-sampled to the same nominal spatial resolution as the clinical images. HiSS water peak height images have stronger T2* weighting without the distortion present in heavily T2*-weighted conventional images, resulting in a high dynamic range (Figs. 7A, 7B, 7C, 7D, 7E, and 7F) that is generally not found in standard clinical T1-weighted high-resolution images. Resolution and anatomic detail are better than what can be obtained in standard images—even at the same nominal spatial resolution—because spectral information is used to eliminate the chemical shift effect that causes blurring at the fat-water interfaces. Blurring due to magnetic susceptibility gradients is also minimized because images are synthesized from only one (in the work presented here, the maximum) frequency component of water resonance and not the whole water resonance, which may be significantly broadened. Finally, superior and more uniform fat suppression is obtained by acquiring and processing high-resolution proton spectral data to discard all nonwater signal.

A novel source of contrast in HiSS images shown here is the information encoded in the structure of the water resonance and the effect of contrast agent on different water signal components (Figs. 7A, 7B, 7C, 7D, 7E, and 7F). This information is generally not available from standard clinical imaging sequences. Earlier work in this laboratory [17-19, 21, 22] and many other studies [23-26] support the hypothesis that the different components in the water signal originate from different water compartments within the voxel. Thus, it may be possible to obtain subvoxel information or contrast (or both) based on spectral information. Because the contrast agent is likely to access the extravascular extracellular, but not the intracellular water compartment, its effects may also facilitate identification of signal from specific microscopic compartments [23, 24, 27, 28]. Our earlier data [18] showed that changes of the sort shown in Figures 7A, 7B, 7C, 7D, 7E, and 7F are well above the noise level and can be analyzed to produce images with clinically useful contrast. Thus, the benefit of HiSS to breast imaging may extend beyond providing higher quality images, to providing a novel MRI contrast that may increase the specificity of breast MRI.

The acquisition time required for data presented here ({approx} 2 min) limits the volume of tissue that can be imaged. However, previous work [16] showed that spectral resolution can be reduced by a factor of 2—without a significant loss in image contrast—cutting the acquisition time in half. Reduced spatial resolution (down to 1.3 x 1.3 mm) would be diagnostically acceptable while providing another factor of 4 in acquisition speed. Acquiring interleaved k-space lines for two slices after a single excitation without changing the TR would result in the reduction of bandwidth, but not spectral resolution, and could reduce the scanning time by another factor of 2. Finally, the implementation of parallel imaging would increase the speed of acquisition by another factor of 2 or more. Implementation of all these changes could speed up the acquisition by a factor of 32 or higher and allow application of HiSS imaging on large volumes of tissue (e.g., covering the whole breast), in thinner slices, or in very short acquisition times. The postprocessing times could also be improved to allow real-time processing through algorithm optimization, using faster programming languages (e.g., C++) and with future faster hardware.

These anticipated improvements in data acquisition and processing make whole breast HiSS scans potentially feasible and valuable. HiSS images generally depict lesions well even before contrast agent administration and could be used to replace T2-weighted precontrast series and high-resolution T1-weighted postcontrast fat-saturated images. A small volume in and around suspicious lesions located in HiSS data could be scanned with high temporal and spatial resolution during contrast medium uptake and washout. This would improve the sensitivity and specificity of dynamic contrast-enhanced MRI scans to rapid tumor blood flow and high capillary permeability. Additional gains in specificity could come from postcontrast HiSS scans to image the effects of contrast agents on the shape of the water line and the resonance frequency in each voxel The precontrast HiSS scans could also be useful for MR-guided percutaneous procedures. Thus, HiSS imaging in combination with dynamic contrast-enhanced MRI is an attractive protocol for breast imaging.

In conclusion, the results demonstrate the feasibility of imaging breast lesions with realistic size distributions using the HiSS sequence in a clinical setting. Although the scanning and processing times are currently long, significant improvements in speed are likely. Even at its current stage of development, HiSS imaging added diagnostic information through excellent margin and internal definition and lesion conspicuity without the use of contrast agents. Our results support further development of HiSS imaging in at least two directions. First, MRI specificity could be improved by developing a new source of subvoxel information based on water resonance structure and changes in the shape of the water signal line due to contrast agents. Second, using spectral information to eliminate all nonwater signal, as well as susceptibility gradient and chemical shift blurring, HiSS imaging could improve the sensitivity of breast MRI by providing superior full-breast-coverage images before contrast agent administration.


References
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
 References
 

  1. Kneeshaw PJ, Turnbull LW, Drew PJ. Current applications and future direction of MR mammography. Br J Cancer2003; 88:4 -10[CrossRef][Medline]
  2. Kinkel K, Vlastos G. MR imaging: breast cancer staging and screening. Semin Surg Oncol 2001;20 : 187-196[CrossRef][Medline]
  3. Esserman L, Wolverton D, Hylton N. Magnetic resonance imaging for primary breast cancer management: current role and new applications. Endocr Relat Cancer 2002;9 : 141-153[Abstract]
  4. Kinkel K, Helbich TH, Esserman LJ, et al. Dynamic high-spatial-resolution MR imaging of suspicious breast lesions: diagnostic criteria and interobserver variability. AJR2000; 175:35 -43[Abstract/Free Full Text]
  5. Kuhl CK, Mielcareck P, Klaschik S, et al. Dynamic breast MR imaging: are signal intensity time course data useful for differential diagnosis of enhancing lesions? Radiology1999; 211:101 -110[Abstract/Free Full Text]
  6. Szabó BK, Aspelin P, Wiberg MK, Boné B. Dynamic MR imaging of the breast: analysis of kinetic and morphologic diagnostic criteria. Acta Radiol 2003;44 : 379-386[CrossRef][Medline]
  7. Wedegärtner U, Bick U, Wörtler K, Rummeny E, Bongartz G. Differentiation between benign and malignant findings on MR-mammography: usefulness of morphological criteria. Eur Radiol2001; 11:1645 -1650[CrossRef][Medline]
  8. Kvistad KA, Rydland J, Vainio J, et al. Breast lesions: evaluation with dynamic contrast-enhanced T1-weighted MR imaging and with T2*-weighted first-pass perfusion MR imaging. Radiology 2000;216 : 545-553[Abstract/Free Full Text]
  9. Sinha S, Sinha U. Functional magnetic resonance of human breast tumors: diffusion and perfusion imaging. Ann N Y Acad Sci 2002; 980:95 -115[Medline]
  10. McKnight AL, Kugel JL, Rossman PJ, Manduca A, Hartmann LC, Ehman RL. MR elastography of breast cancer: preliminary results. AJR 2002; 178:1411 -1417[Abstract/Free Full Text]
  11. Yuen S, Yamada K, Kinosada Y, et al. Equivalent cross-relaxation rate imaging of breast cancer. J Magn Reson Imaging2004; 20:56 -65[CrossRef][Medline]
  12. Larkman DJ, deSouza NM, Bydder M, Hajnal JV. An investigation into the use of sensitivity-encoded techniques to increase temporal resolution in dynamic contrast-enhanced breast imaging. J Magn Reson Imaging 2001; 14:329 -335[CrossRef][Medline]
  13. Fletcher DW, Haselgrove JC, Bolinger L. High-resolution imaging using Hadamard encoding. Magn Reson Imaging1999; 17:1457 -1468[CrossRef][Medline]
  14. Alterson R, Plewes DB. Bilateral symmetry analysis of breast MRI. Phys Med Biol 2003;48 : 3431-3443[CrossRef][Medline]
  15. Kuperman VY, Alley MT. Differentiation between the effects of T1 and T2* shortening in contrast-enhanced MRI of the breast. J Magn Reson Imaging 1999;9 : 172-176[CrossRef][Medline]
  16. Medved M, Du W, Zamora MA, et al. The effect of varying spectral resolution on the quality of high spectral and spatial resolution magnetic resonance images of the breast. J Magn Reson Imaging2003; 18:442 -448[CrossRef][Medline]
  17. Du W, Du YP, Bick U, et al. Breast MR imaging with high spectral and spatial resolutions: preliminary experience. Radiology 2002;224 : 577-585[Abstract/Free Full Text]
  18. Medved M, Newstead GM, Fan X, et al. Fourier components of inhomogeneously broadened water resonances in breast: a new source of MRI contrast. Magn Reson Med 2004;52 : 193-196[CrossRef][Medline]
  19. Kovar DA, Al-Hallaq HA, Zamora MA, River JN, Karczmar GS. Fast spectroscopic imaging of water and fat resonances to improve the quality of MR images. Acad Radiol 1998;5 : 269-275[CrossRef][Medline]
  20. Mansfield P. Spatial mapping of the chemical shift in NMR. Magn Reson Med 1984;1 : 370-386[Medline]
  21. Karczmar GS, Fan X, Al-Hallaq H, et al. Functional and anatomic imaging of tumor vasculature: high-resolution MR spectroscopic imaging combined with a superparamagnetic contrast agent. Acad Radiol 2002; 9[suppl 1]:S115 -S118[Medline]
  22. Karczmar GS, Fan X, Al-Hallaq HA, et al. Uptake of a superparamagnetic contrast agent imaged by MR with high spectral and spatial resolution. Magn Reson Med 2000;43 : 633-639[CrossRef][Medline]
  23. Naritomi H, Kanashiro M, Sasaki M, Kuribayashi Y, Sawada T. In vivo measurements of intra- and extracellular Na+ and water in the brain and muscle by nuclear magnetic resonance spectroscopy with shift reagent. Biophys J 1987;52 : 611-616[Medline]
  24. Zhong K, Li X, Shachar-Hill Y, Picart F, Wishnia A, Springer CS Jr. Magnetic susceptibility shift selected imaging (MESSI) and localized (1)H(2)O spectroscopy in living plant tissues. NMR Biomed2000; 13:392 -397[CrossRef][Medline]
  25. Gambarota G, Cairns BE, Berde CB, Mulkern RV. Osmotic effects on the T2 relaxation decay of in vivo muscle. Magn Reson Med 2001; 46:592 -599[CrossRef][Medline]
  26. Peled S, Cory DG, Raymond SA, Kirschner DA, Jolesz FA. Water diffusion, T(2), and compartmentation in frog sciatic nerve. Magn Reson Med 1999; 42:911 -918[CrossRef][Medline]
  27. Brindle KM, Brown FF, Campbell ID, Grathwohl C, Kuchel PW. Application of spin-echo nuclear magnetic resonance to whole-cell systems: membrane transport. Biochem J 1979;180 : 37-44[Medline]
  28. Quirk JD, Bretthorst GL, Duong TQ, et al. Equilibrium water exchange between the intra- and extracellular spaces of mammalian brain. Magn Reson Med 2003;50 : 493-499[CrossRef][Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 radtechHome page
T. G. ODLE
Breast MR
Radiol. Technol., September 1, 2006; 78(1): 45M - 66M.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Medved, M.
Right arrow Articles by Karczmar, G. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Medved, M.
Right arrow Articles by Karczmar, G. S.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME