DOI: 10.1148/rg.265055176
RadioGraphics 2006;26:1469-1484
© RSNA, 2006
How to Optimize Clinical Breast MR Imaging Practices and Techniques on Your 1.5-T System1
Dana R. Rausch, MD and
R. Edward Hendrick, PhD
1 From the Department of Radiology, Mount Sinai Medical Center, Box 1234, 1 Gustave L. Levy Place, New York, NY 10029 (D.R.R.); and Department of Radiology, Lynn Sage Comprehensive Breast Center, Northwestern University Feinberg School of Medicine, Chicago, Ill (R.E.H.). Recipient of a Certificate of Merit award for an education exhibit at the 2004 RSNA Annual Meeting. Received September 21, 2005; revision requested January 4, 2006, and received January 31; accepted February 6. D.R.R. has been an educational speaker for Suros Surgical Systems, and R.E.H. is an educational speaker for GE Healthcare.
Address correspondence to D.R.R. (e-mail: danarausch{at}hotmail.com).
 |
Abstract
|
|---|
Magnetic resonance (MR) imaging, when used in conjunction with mammography and ultrasonography, can be a powerful tool for breast imaging. There are various clinical scenarios in which MR imaging may provide key information that leads to an alteration in treatment plans (eg, by demonstrating features that were occult at physical examination or conventional imaging). Although many benign and malignant entities enhance at contrast material–enhanced breast MR imaging, the morphologic characteristics and kinetic profiles of lesions help narrow the differential diagnosis. To optimize the quality of the morphologic and kinetic information yielded by breast MR imaging, the radiologist must attend to various practical and technical prerequisites: A bilateral breast coil should be used with prone positioning of the patient. An MR imaging system with a high-field-strength magnet is needed, and the magnetic field must be homogeneous across the field of view, which should include both breasts. A T2-weighted sequence should be applied first to identify any cysts and should be followed by three-dimensional imaging with a T1-weighted spoiled gradient-echo sequence after the intravenous administration of a gadolinium chelate. To minimize artifacts, a direction other than the anterior-posterior direction should be selected for phase encoding. To suppress the signal from fat, a frequency-selective pulse should be applied during imaging, or the unenhanced MR imaging data should be subtracted from the contrast-enhanced MR imaging data during postprocessing. The imaging section thickness should be 3 mm or less, the pixel size should be less than 1 mm in each in-plane direction, and the total acquisition time should be less than 2 minutes.
A patient questionnaire to supplement this article is available at http://radiographics.rsnajnls.org/cgi/content/full/26/5/1469/DC1
© RSNA, 2006
|
LEARNING OBJECTIVES
|
|---|
After reading this article and taking the test, the reader will be able to:- Identify the indications and contraindications for breast MR imaging.
- Discuss the technical and practical requirements for optimizing breast MR imaging.
- Describe a comprehensive approach to breast MR image interpretation.
|
Introduction
|
|---|
In the early to middle 1980s, the use of magnetic resonance (MR) imaging to differentiate malignant breast lesions from normal breast tissue and benign breast lesions on the basis of inherent tissue longitudinal relaxation times (T1), transverse relaxation times (T2), and hydrogen spin densities was investigated (1–4). Malignant breast lesions were found to have higher T1 and T2 values than normal breast tissue but shorter T1 and T2 values than some benign breast lesions, such as fibroadenomas. The overlap in T1 and T2 values between benign and malignant lesions previously discouraged the use of unenhanced breast MR imaging for cancer detection and diagnosis. However, the application of unenhanced breast MR imaging was recently revisited, with the use of modern imaging techniques and a multiparametric approach, to differentiate benign from malignant lesions (5,6). Despite a reasonable degree of success in lesion differentiation with this method, it is generally considered that MR imaging must be performed both before and after the administration of a paramagnetic contrast agent such as gadopentetate dimeglumine to achieve high sensitivity for the detection of breast cancer.
For more than 15 years, contrast material–enhanced breast MR imaging has been investigated as a means of cancer detection and diagnosis (7–14). The use of gadopentetate dimeglumine (Magnevist; Berlex, Wayne, NJ, and Schering, Munich, Germany) increases the sensitivity of breast MR imaging for detection of breast cancer. Sensitivities as high as 83%–100% have been reported, with specificities of 29%–100% depending on the technique and diagnostic criteria used (10–12).
In the early 1990s, two competing approaches to contrast-enhanced breast MR imaging emerged. One approach was aimed at achieving high spatial resolution for clear depiction of the morphologic features of enhanced lesions, including their shape, edge characteristics, and internal structure (15,16). The other approach was devised to collect dynamic information about contrast material uptake and washout for improved specificity, usually at the expense of high spatial resolution (17,18).
More recent developments in MR gradient systems and pulse sequences allow the simultaneous achievement of high spatial resolution and adequate temporal resolution.
The article surveys the technical and practical requirements for obtaining high-quality breast MR images and reviews clinical considerations and pitfalls of breast MR image interpretation.
|
Clinical Prerequisites
|
|---|
A radiologist should be assigned the role of gatekeeper and the responsibility for approving all breast MR imaging examinations. Ideally, the gatekeeper should have specific training in breast imaging and should be adept at applying MR imaging protocols and techniques as well as those of conventional imaging modalities such as mammography and ultrasonography (US). Knowledge and experience in the interpretation of mammograms and US images enables the comparison of mammographic and US findings with findings on breast MR images. The gatekeeper should educate referring surgeons and medical and radiation oncologists about appropriate clinical indications for breast MR imaging and should ensure that contrast-enhanced imaging is performed between days 7 and 14 of the menstrual cycle (day 1 being the day on which menstruation begins) in pre-menopausal women.
Proposed Indications
A recently published American College of Radiology practice guideline specifies the proposed indications for breast MR imaging (Table 1) (19). All of the indications refer to contrast-enhanced breast MR imaging, except in the case of a suspected silicone breast implant rupture.
Inconclusive Findings of Breast Cancer.—
MR imaging may be helpful for lesion characterization when findings at physical examination and at thorough and tailored examinations with conventional imaging modalities are suggestive of breast cancer but are inconclusive (eg, a finding such as asymmetric density or a lesion depicted on only one mammographic view). In patients who have undergone surgery for a malignancy, MR imaging may be used to distinguish scar tissue from tumor recurrence. The modality can be particularly useful for examining breasts that appear dense on mammograms.
Neoadjuvant Chemotherapy.—
In patients with locally advanced breast cancer, MR imaging may be performed before, during, or after a course of chemotherapy to evaluate the tumor response and the extent of residual disease before surgical intervention. If breast-conserving therapy (ie, lumpectomy and subsequent radiation therapy) is planned, a tissue marking clip should be placed within the malignancy by the radiologist in case the mammographic, US, and clinical findings are no longer apparent at imaging after treatment. Accurate measurement of the residual tumor may be difficult, as the tumor response may not be manifested in a pattern of uniform concentric changes. Rather, islands of tumor cells typically are intermixed with necrotic and nonenhancing fibrotic tissue (20). Commercially available computer software programs such as DynaCAD (In Vivo, Orlando, Fla) and CADstream (Confirma, Kirkland, Wash) enable the calculation of tumor volumes for quantification of the response to chemotherapy.
Infiltrating Carcinoma.—
The diagnosis of infiltrating lobular carcinoma may be difficult on the basis of conventional breast imaging and physical examination because of its linear growth pattern. Because of the low vascular and cellular density of lesions of this type, there are specific limitations to their diagnosis on the basis of MR imaging features, as well. MR imaging may be useful for evaluating the size, extent, and possible multicentricity (multiple tumor sites in a different quadrant than, or at a distance of more than 5 cm from, the index cancer) or multifocality (multiple tumor sites in the same quadrant as, or within 5 cm of, the index cancer) of infiltrating carcinoma. However,
in patients who are undergoing MR imaging to determine whether multifocal or multicentric disease is present, it should not be assumed without histologic confirmation that all enhanced lesions are malignant, as such an assumption may lead to treatment with mastectomy when lumpectomy is more appropriate. The ability to provide MR image-guided breast biopsy and localization of lesions depicted only on MR images is crucial.
Axillary Adenopathy.—
When axillary adenopathy is present but the findings at mammography, US, and physical examination are negative for primary malignancy, MR images may depict the site of primary malignancy in the breast. Thus, MR imaging may facilitate management with breast-conserving therapy instead of mastectomy, which traditionally has been performed when the primary site is unknown.
Postoperative Reconstruction.—
In patients who have undergone complete or partial breast reconstruction with autologous myocutaneous flaps (ie, from the rectus abdominis, latissimus dorsi, or gluteus maximus muscles) or implants, MR imaging may be helpful for detecting a tumor recurrence when the findings at conventional imaging are inconclusive.
Breast Augmentation.—
The detection of breast cancer at mammography can be challenging in patients with silicone or other implants or free silicone injections, and contrast-enhanced breast MR imaging may be used as a supplement to routine conventional imaging in those patients.
Suspected Invasion Deep to Pectoral Fascia.—
MR imaging is most useful in the preoperative evaluation of patients in whom breast cancer is suspected to have invaded the pectoralis major, serratus anterior, or intercostal muscle.
Possible Contralateral Breast Lesion.—
In patients in whom a malignancy has been found in one breast at mammography or US, MR imaging may depict a previously occult malignancy in the contralateral breast.
Previous Lumpectomy.—
In patients who underwent lumpectomy without preoperative MR imaging and those with close or positive surgical margins after lumpectomy, the extent of residual disease and possible multifocality or multicentricity demonstrated on MR images may help determine whether repeat excision or mastectomy is appropriate. Visualization of residual disease may be more difficult with mammography because of scar tissue related to lumpectomy.
High Individual Cancer Risk.—
Patients at increased risk for breast cancer because of a genetic predisposition or previous mediastinal irradiation may benefit from supplementary screening with breast MR imaging in addition to mammography and US.
Suspicion of Cancer Recurrence.—
MR imaging may be helpful for the assessment of patients who have undergone breast-conserving therapy and subsequent radiation therapy, as conventional imaging and clinical evaluation may be difficult and inconclusive, especially in women with dense breasts.
Relative Contraindications
Breast MR imaging does not replace conventional imaging (ie, mammography and US) and physical examination. There is no role for breast MR imaging in the differentiation of benign from malignant findings at mammography, US, or clinical examination and no justification for the addition of breast MR imaging to the modalities used for routine screening in non-high-risk asymptomatic women. Such uses of MR imaging would entail an inappropriate utilization of resources and might increase patient anxiety unnecessarily. Moreover, a negative breast MR imaging examination does not exclude the presence of a malignancy or preclude the appropriate management of an otherwise suspicious finding.
Other Pertinent Clinical Information
In addition to the standard screening questionnaire to identify the presence of contraindications to MR imaging (eg, cardiac pacemakers, cochlear implants, and ferromagnetic aneurysm clips), a clinical questionnaire should be completed by the patient before all breast MR imaging studies (see Appendix E1 at radiographics.rsnajnls.org/cgi /content/full/26/5/000/DC1). The completed questionnaire should be reviewed with the patient by a registered nurse, medical assistant, or trained MR imaging technologist. The information it contains is critical because the management of an incidental enhancing lesion found at MR imaging differs for a woman with a previously diagnosed high-risk lesion identified at breast biopsy (eg, lobular carcinoma in situ, atypia, or radial scar) or a previous breast malignancy, as well as for a woman who has a genetic predisposition to breast cancer or who is undergoing chemopreventive therapy with a medication such as tamoxifen citrate.
|
Technical Requirements
|
|---|
The technical requirements for breast MR imaging are listed in Table 2 and described in detail in the following subsections.
Magnetic Field Strength and Homogeneity
There is a linear relationship between magnetic field strength (B0) and signal-to-noise ratio (SNR). With greater B0, the SNR is higher, and images with higher spatial resolution can be obtained in a relatively short acquisition time if appropriate pulse sequences are used. The magnetic field should be homogeneous across the entire field of view, which includes both breasts. With low to intermediate field strengths (ie, less than 1.0 T), a lack of adequate homogeneity prevents chemically selective fat suppression and may compromise image quality.
Use of a Prone-positioning Bilateral Breast Coil
Mild compression applied to the breast in the lateral-to-medial direction decreases the amount of tissue to be imaged in that direction and thereby decreases the image acquisition time, an effect that is particularly useful for imaging in the axial and sagittal planes. Compression also decreases patient motion during each sequence and between sequences and thus enables the avoidance of signal misregistration on subtracted images. Compression should be applied gently, as firm compression may inhibit contrast agent uptake in the breasts.
Bilateral imaging allows the assessment of breast symmetry, as at mammography. It increases the conspicuity of asymmetric diffuse enhancement, which may be difficult to identify on unilateral views (Fig 1). Bilateral imaging diminishes the number of false-positive findings caused by the spurious enhancement that typically occurs in the hormonally active breast of a premenopausal woman or a woman receiving hormone replacement therapy. It also may aid in the detection of a synchronous and otherwise occult cancer in the contralateral breast in approximately 3%–5% of women with a diagnosis of breast cancer (21–23).
View larger version (84K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1. Axial contrast-enhanced T1-weighted fat-suppressed MR image demonstrates asymmetric diffuse heterogeneous enhancement in the left breast in a patient with infiltrating ductal carcinoma.
|
|
Unilateral breast imaging also has technical disadvantages. For example, images acquired in the axial or the coronal plane in one breast may be degraded by wraparound artifacts from the contralateral breast, particularly when the phase-encoding direction is from left to right. Wraparound artifacts, accompanied by a decrease in the SNR, also may occur when a bilateral breast coil is used and the receiver coil for one breast is left on during imaging of the other breast. A significant wraparound artifact occurs when the field of view is smaller than the patient and the signal from tissue outside the selected field of view wraps around or aliases into the acquired image. The artifacts occur primarily in the phase-encoding direction. Possible solutions are to enlarge the field of view or to apply the option of no phase wrap, which increases the number of phase-encoding views acquired in the phase-encoding direction but does not present them all in the acquired image.
Multichannel coils provide a higher SNR and more uniform image intensity across both breasts. Both the four-channel bilateral breast coil and the newer seven- and 12-channel bilateral coils are compatible with current parallel imaging techniques. Such techniques permit a reduction in acquisition time by using the sensitivity profiles of the individual channels to simultaneously acquire different sections of the image, or different views in k-space, instead of using all the channels to acquire the same image data (24). Compared with conventional imaging techniques, parallel imaging can halve the image acquisition time through simultaneous imaging of both breasts and can reduce effects such as wraparound, albeit at the expense of a 41% decrease in SNR, if all other imaging parameters are the same. Combining the partial data sets acquired with the individual channels in parallel imaging requires the use of image reconstruction techniques that are more complex than those of two-dimensional or three-dimensional (3D) imaging.
Unenhanced T2-weighted Imaging to Identify Cysts
Because of their high water content, cysts have longer T1 and T2, and higher spin density values than do normal breast tissues, cancers, and benign lesions. This characteristic makes cysts appear bright on MR images acquired with T2-weighted, spin-echo, fast spin-echo, and short inversion time inversion recovery sequences (Fig 2).
View larger version (100K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. (a) Axial MR image obtained with a short inversion time inversion recovery sequence shows several areas of uniform high signal intensity in the breast, a finding that signifies one or more cysts. (b) Axial US image at a similar level in the breast demonstrates an oval, macrolobulated, circumscribed, horizontally oriented mass with posterior acoustic enhancement. This finding corresponds to that in a and is compatible with the diagnosis of a benign simple cyst.
|
|
View larger version (149K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. (a) Axial MR image obtained with a short inversion time inversion recovery sequence shows several areas of uniform high signal intensity in the breast, a finding that signifies one or more cysts. (b) Axial US image at a similar level in the breast demonstrates an oval, macrolobulated, circumscribed, horizontally oriented mass with posterior acoustic enhancement. This finding corresponds to that in a and is compatible with the diagnosis of a benign simple cyst.
|
|
Contrast-enhanced T1-weighted Imaging
For contrast-enhanced T1-weighted imaging, a spoiled gradient-echo sequence is applied with a short repetition time, a very short echo time, and a moderate flip angle (25). Since more signal is obtained in each measurement with volume acquisition in 3D Fourier transform sequences, that technique has an SNR advantage over two-dimensional Fourier transform technique, in which the signal is acquired from a single plane at a time. One sequence is applied before the contrast agent is administered, and four or more contrast-enhanced acquisitions follow; contrast-enhanced imaging continues until approximately 7 minutes after the contrast agent injection. The ideal temporal resolution is 1–2 minutes per acquisition, so that the uptake and washout of the contrast agent, which in the most rapidly enhancing lesions typically occurs within 90–120 seconds after injection, can be measured accurately.
Selection of the Phase-encoding Direction
Cardiac and respiratory motion may lead to the propagation of artifacts across the breasts in the in-plane phase-encoding direction. To minimize such effects, the in-plane phase-encoding direction should never be anterior-posterior. The frequency-encoding direction should be anterior-posterior. For sagittal and coronal imaging, the phase-encoding direction should be superior-inferior. For axial imaging, the phase-encoding direction should be left-right (Fig 3).
View larger version (106K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. (a) Axial contrast-enhanced T1-weighted MR image acquired with improper selection of the anterior-posterior direction as the phase-encoding direction shows a resultant cardiac motion–related artifact propagated across the breasts in a vertical direction (arrows). (b) Axial T1-weighted image acquired in another patient with the left-right direction properly selected as the phase-encoding direction (arrows) provides better depiction of the breasts.
|
|
View larger version (87K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. (a) Axial contrast-enhanced T1-weighted MR image acquired with improper selection of the anterior-posterior direction as the phase-encoding direction shows a resultant cardiac motion–related artifact propagated across the breasts in a vertical direction (arrows). (b) Axial T1-weighted image acquired in another patient with the left-right direction properly selected as the phase-encoding direction (arrows) provides better depiction of the breasts.
|
|
Use of a Gadolinium-based Contrast Agent
Elemental gadolinium is toxic, but chelation (in formulations such as gadopentetate dimeglumine) renders it nontoxic. The paramagnetic properties of the gadolinium atom are due to three unpaired electrons in its outer electron shell. The magnetic dipole moments of the unpaired electrons cause the gadolinium atoms to align with an externally applied magnetic field and to add to the total magnetization. However, the atom exhibits no magnetic effects when it is not exposed to a magnetic field.
Benign and malignant breast lesions may appear similar on unenhanced MR images, but the sensitivity of breast MR imaging for detection of malignancies is improved with the acquisition of images both before and after the injection of a paramagnetic contrast agent such as a gadolinium chelate (eg, gadopentetate dimeglumine, Magnevist, Berlex, Wayne, NJ; gadodiamide, Omniscan, Nycomed Amersham, Oslo, Norway; or gadoteridol, ProHance, Bracco Diagnostics, Princeton, NJ) (26). The paramagnetic properties of the contrast agent result in decreased T1, T2, and T2* relaxation times. Since the decrease in relaxation time is greatest for T1, T1-weighted sequences have been used to image gadolinium chelates (27). After its intravenous injection, the contrast agent circulates in the bloodstream and is distributed to extracellular spaces where blood vessels are "leaky" because of malignant angiogenesis. Gadolinium chelates serve as positive contrast agents, enhancing the signal intensity of lesions, vascular structures, and extracellular spaces in the vicinity of vascular lesions. Shorter T1 values make tissues appear brighter on T1-weighted images acquired after administration of the agent. After circulating through the bloodstream, the gadolinium chelate is collected in the kidneys through glomerular filtration. Gadolinium chelates have an elimination half-life of a few hours (26).
For breast MR imaging in the clinical setting, a gadolinium chelate is injected intravenously at a dose of 0.1–0.2 mmol per kilogram of body weight. The gadolinium chelate injection is followed with a 20-mL saline flush to ensure that the contrast agent is cleared from the intravenous tubing and is in circulation. The injections are administered with a power injector at a rate of 1–2 mL/sec to achieve consistency in the timing of contrast enhancement. The administration of a consistent contrast agent dose according to body weight and of a fixed amount of saline at a consistent injection rate is crucial to obtain consistent image quality at breast MR imaging.
Homogeneous Fat Suppression
Uniform fat suppression minimizes the signal from fat, which otherwise might decrease the conspicuity of a contrast-enhanced lesion on T1-weighted images. Fat suppression may be achieved with a frequency-selective saturation pulse at the time of image acquisition or with subtraction sequences during postprocessing.
Chemically selective suppression of the signal from fat at T1-weighted 3D Fourier transform imaging is achieved by applying a fat frequency–selective 90° saturation pulse to the entire volume of breast within the coil. When fat suppression is inhomogeneous, the center frequency should be changed to that of fat, which at a field strength of 1.5 T is 220 Hz below that of water, and the image acquisition should be repeated until the fat appears dark (Fig 4). Radiologists should be aware that the inadvertent application of water signal suppression could result in saturation of the signal from gadolinium chelates and lead to diagnostic errors.
If a fat-saturation pulse was not used in image acquisition, a similar effect may be obtained by subtracting the unenhanced image data from the sequential contrast-enhanced image data on a pixel-by-pixel and image-by-image basis. However, if this method is used, patient motion between acquisitions may cause signal misregistrations that simulate contrast-enhanced lesions on the resultant images and, thus, lead to impaired image interpretation (Fig 5). Such occurrences could result in errors in the diagnosis of small high-signal-intensity contrast-enhanced lesions.
View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Fat suppression by means of image subtraction. (a, b) Sagittal T1-weighted breast MR images acquired before (a) and after (b) contrast agent injection and without a fat-saturation pulse. (c) Subtraction image of the same section as in a and b demonstrates an enhanced mass and several artifacts that mimic areas of contrast agent uptake. The artifacts were caused by signal misregistration between images acquired before and after contrast agent administration.
|
|
View larger version (101K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Fat suppression by means of image subtraction. (a, b) Sagittal T1-weighted breast MR images acquired before (a) and after (b) contrast agent injection and without a fat-saturation pulse. (c) Subtraction image of the same section as in a and b demonstrates an enhanced mass and several artifacts that mimic areas of contrast agent uptake. The artifacts were caused by signal misregistration between images acquired before and after contrast agent administration.
|
|
View larger version (68K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5c. Fat suppression by means of image subtraction. (a, b) Sagittal T1-weighted breast MR images acquired before (a) and after (b) contrast agent injection and without a fat-saturation pulse. (c) Subtraction image of the same section as in a and b demonstrates an enhanced mass and several artifacts that mimic areas of contrast agent uptake. The artifacts were caused by signal misregistration between images acquired before and after contrast agent administration.
|
|
Thin-Section Acquisitions
Section thickness is the factor that determines how small a breast lesion can be identified; thus, the use of thinner sections increases the sensitivity of breast MR imaging. For example, to depict a 3-mm lesion, a section thickness of 3 mm or less must be used. Even at this section thickness, a low-contrast lesion may be difficult to detect because of partial volume averaging in the presence of image noise (Fig 6).
View larger version (153K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Effect of section thickness on tissue visibility and image quality. Breast MR images obtained with 1-mm (a), 2-mm (b), 3-mm (c), and 4-mm (d) section thicknesses show that as section thickness increases, SNR also increases, but so does the severity of the partial volume artifact (blurring of tissue margins).
|
|
View larger version (139K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Effect of section thickness on tissue visibility and image quality. Breast MR images obtained with 1-mm (a), 2-mm (b), 3-mm (c), and 4-mm (d) section thicknesses show that as section thickness increases, SNR also increases, but so does the severity of the partial volume artifact (blurring of tissue margins).
|
|
View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6c. Effect of section thickness on tissue visibility and image quality. Breast MR images obtained with 1-mm (a), 2-mm (b), 3-mm (c), and 4-mm (d) section thicknesses show that as section thickness increases, SNR also increases, but so does the severity of the partial volume artifact (blurring of tissue margins).
|
|
View larger version (141K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6d. Effect of section thickness on tissue visibility and image quality. Breast MR images obtained with 1-mm (a), 2-mm (b), 3-mm (c), and 4-mm (d) section thicknesses show that as section thickness increases, SNR also increases, but so does the severity of the partial volume artifact (blurring of tissue margins).
|
|
Small Pixel Size
Smaller pixel sizes provide increased definition of the margins, shape, and internal features of a lesion and improve the depiction of irregular margins and spicules. At bilateral imaging in the sagittal plane, a pixel size of 0.78 x 0.78 mm can be achieved by using a field of view of 20 cm or less with a 256 x 256 matrix. At axial or coronal imaging, a pixel size of 0.71 x 0.71 mm can be obtained with a field of view of 32 cm and a 448 x 448 matrix. A pixel size of less than 1.0 mm in each in-plane direction is necessary for optimal spatial resolution at breast MR imaging.
Adequate Temporal Resolution
The required temporal resolution is determined by the time course of contrast agent uptake. Peak contrast enhancement in a malignant lesion typically occurs between 90 and 180 seconds after injection of the contrast agent, so a temporal resolution of less than 2 minutes is crucial for accurate depiction of the kinetics of lesion contrast enhancement (Fig 7). At bilateral breast imaging, both breasts must be imaged within that length of time. Typically, unenhanced and contrast-enhanced sequences are applied with identical imaging parameters so that an image obtained with subtraction of the unenhanced image data from the sequential contrast-enhanced image data will reflect temporal changes only in the extent of contrast agent uptake. Imaging for approximately 6–7 minutes after contrast agent injection is sufficient to determine the shape of the time-enhancement curve.
View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7. Typical time-enhancement curves. Type III (gray curve) and type II (black curve) lesion enhancement can be accurately distinguished at 11/4 minutes after contrast material administration (vertical dashed line), but the distinction is lost with delayed imaging at 4 minutes (vertical solid line).
|
|
In performing breast MR imaging, the radiologist must balance a concern for optimizing spatial and temporal resolution while maintaining an adequate SNR. Higher spatial resolution requires thinner sections, which result in a higher sensitivity for the detection of breast cancer by better depicting lesion morphology (ie, mass margins, shape and internal enhancement characteristics), especially for more diffuse lesions such as ductal carcinoma in situ. However, increasing spatial resolution results in decreased per-pixel SNR and may compromise temporal resolution. Higher temporal resolution requires rapid sequential imaging, which often has been achieved at the expense of greater section thickness and larger in-plane pixel size. Newer pulse sequences that take advantage of the improved gradient strength and shorter rise times of current MR systems allow for simultaneous image acquisition with high spatial resolution, adequate temporal resolution, and high SNR. It is now possible to perform 3D acquisitions simultaneously in both breasts with submillimeter in-plane pixel size, 1-mm section thickness, 1-minute temporal resolution, and high SNR per pixel. With nearly isotropic 3D acquisitions, the primary plane of acquisition is less important because the acquired data can be reformatted in any plane, or subtracted images can be displayed as maximum intensity projections, without a significant loss of spatial resolution.
|
Practical Requirements
|
|---|
Instruction and Positioning of the Patient
The MR imaging technologist should inform the patient of the approximate duration of the image acquisition. The importance of remaining still during image acquisition should be emphasized. The patient should know that she can communicate with the technologist via the intercom.
The MR imaging technologist should be trained in the appropriate positioning of the patient and of the patient’s breasts within the coil. Every effort should be made to ensure that the patient is comfortable, as this will improve compliance and prevent motion. Her arms should be positioned and stabilized so that they are not lateral to breast tissue, to prevent wraparound artifacts when the left-right phase-encoding direction is correctly selected for axial imaging. Her head should be supported by a head holder or pillow to avoid wraparound artifacts when the superior-inferior phase-encoding direction is correctly chosen for sagittal imaging. A pillow should be placed under the patient’s legs to help her tolerate the prone position.
Orientation of Image Acquisition
It is most appropriate to acquire unenhanced and contrast-enhanced images in the sagittal and axial planes, as these orientations allow better visualization of the ductal system and easier correlation with mammograms. Three-dimensional data sets with nearly isotropic voxels (ie, voxels of the same size in the section thickness direction and in each in-plane direction) may be acquired in any plane, since they can be reconstructed with or without postprocessing (eg, subtraction) in any plane, without a significant loss of spatial resolution.
|
Image Display and Interpretation
|
|---|
Multiplanar reformatted images can be reconstructed from any 3D data set and in any plane to depict the spatial relationship of findings on images in other orientations. Maximum intensity projections may be reconstructed from subtracted 3D data sets to demonstrate the areas of highest signal intensity in the breast, such as arteries, veins, lymph nodes, and contrast-enhanced lesions (Fig 8). In addition, supplemental breast MR imaging software can be used to obtain angiogenesis maps and time-enhancement curves and to perform lesion volume calculations.
View larger version (60K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8a. T1-weighted fat-suppressed breast MR images. (a) Axial image obtained in a single bilateral breast section before contrast agent injection. (b) Axial contrast-enhanced image in the same section as a demonstrates segmentally distributed areas of clumped heterogeneous enhancement in the left breast (arrow). (c) Axial image obtained in another bilateral breast section before contrast agent injection. (d) Axial contrast-enhanced image in the same section as c demonstrates a lobulated, heterogeneously enhanced, irregularly marginated mass (arrow). (e) Maximum intensity projection of the entire set of axial images shows both enhanced lesions in the left breast (arrows).
|
|
View larger version (61K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8b. T1-weighted fat-suppressed breast MR images. (a) Axial image obtained in a single bilateral breast section before contrast agent injection. (b) Axial contrast-enhanced image in the same section as a demonstrates segmentally distributed areas of clumped heterogeneous enhancement in the left breast (arrow). (c) Axial image obtained in another bilateral breast section before contrast agent injection. (d) Axial contrast-enhanced image in the same section as c demonstrates a lobulated, heterogeneously enhanced, irregularly marginated mass (arrow). (e) Maximum intensity projection of the entire set of axial images shows both enhanced lesions in the left breast (arrows).
|
|
View larger version (56K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8c. T1-weighted fat-suppressed breast MR images. (a) Axial image obtained in a single bilateral breast section before contrast agent injection. (b) Axial contrast-enhanced image in the same section as a demonstrates segmentally distributed areas of clumped heterogeneous enhancement in the left breast (arrow). (c) Axial image obtained in another bilateral breast section before contrast agent injection. (d) Axial contrast-enhanced image in the same section as c demonstrates a lobulated, heterogeneously enhanced, irregularly marginated mass (arrow). (e) Maximum intensity projection of the entire set of axial images shows both enhanced lesions in the left breast (arrows).
|
|
View larger version (57K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8d. T1-weighted fat-suppressed breast MR images. (a) Axial image obtained in a single bilateral breast section before contrast agent injection. (b) Axial contrast-enhanced image in the same section as a demonstrates segmentally distributed areas of clumped heterogeneous enhancement in the left breast (arrow). (c) Axial image obtained in another bilateral breast section before contrast agent injection. (d) Axial contrast-enhanced image in the same section as c demonstrates a lobulated, heterogeneously enhanced, irregularly marginated mass (arrow). (e) Maximum intensity projection of the entire set of axial images shows both enhanced lesions in the left breast (arrows).
|
|
View larger version (54K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8e. T1-weighted fat-suppressed breast MR images. (a) Axial image obtained in a single bilateral breast section before contrast agent injection. (b) Axial contrast-enhanced image in the same section as a demonstrates segmentally distributed areas of clumped heterogeneous enhancement in the left breast (arrow). (c) Axial image obtained in another bilateral breast section before contrast agent injection. (d) Axial contrast-enhanced image in the same section as c demonstrates a lobulated, heterogeneously enhanced, irregularly marginated mass (arrow). (e) Maximum intensity projection of the entire set of axial images shows both enhanced lesions in the left breast (arrows).
|
|
Breast MR images should be displayed and reviewed on a picture archiving and communication system workstation to allow optimization of windowing and comparisons with digitally stored images from previous breast imaging examinations. Recent mammograms and US images should be available for review and correlation of findings. Correlation is especially important in cases in which microcalcifications are depicted at mammography, since MR imaging has a lower sensitivity for the detection of ductal carcinoma in situ than for that of other breast cancers.
Spiculated and irregularly marginated masses and areas of ductal and segmentally distributed enhancement on MR images, as on mammograms, are signs suggestive of malignancy regardless of the contrast enhancement kinetics, provided that enhancement is sufficient (Figs 9, 10).
View larger version (134K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9. Sagittal contrast-enhanced T1-weighted fat-suppressed image shows an irregular, heterogeneously enhanced mass with a thick peripheral rim, a finding that represents an invasive ductal carcinoma. Note the superficial enhanced vessels in the superior portion of the breast.
|
|
View larger version (152K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Ductal carcinoma in situ. (a) Sagittal contrast-enhanced T1-weighted fat-suppressed image shows segmentally distributed areas of clumped enhancement in the inferior portion of the breast, a typical finding of ductal carcinoma in situ. (b) Sagittal contrast-enhanced T1-weighted fat-suppressed image demonstrates areas of clumped enhancement in a linear configuration in the breast of another patient who presented with a malignant axillary lymph node. Findings at mammography and US were unremarkable. Results of pathologic analysis helped confirm the diagnosis of ductal carcinoma in situ in both cases.
|
|
View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Ductal carcinoma in situ. (a) Sagittal contrast-enhanced T1-weighted fat-suppressed image shows segmentally distributed areas of clumped enhancement in the inferior portion of the breast, a typical finding of ductal carcinoma in situ. (b) Sagittal contrast-enhanced T1-weighted fat-suppressed image demonstrates areas of clumped enhancement in a linear configuration in the breast of another patient who presented with a malignant axillary lymph node. Findings at mammography and US were unremarkable. Results of pathologic analysis helped confirm the diagnosis of ductal carcinoma in situ in both cases.
|
|
Time-enhancement curves are used to analyze enhanced lesions by displaying signal intensity values versus time, throughout the course of contrast agent uptake. On the original nonsubtracted images, a region of interest (ROI) should be placed over the area with the most robust enhancement or the most suspicious washout curve pattern, while excluding obvious vessels. The size of the ROI should be greater than three pixels (28). Depending on the injection rate and contrast agent dose, most institutions set the threshold level of enhancement at a greater than 50% increase in signal intensity beyond that before contrast agent injection. To avoid spurious time-enhancement curves and resultant diagnostic errors, the radiologist should be sure, when positioning an ROI, that the patient has not moved between image acquisitions and that the ROI on all contrast-enhanced images includes the lesion. When time-enhancement curves are obtained from several ROIs, the curves that are most worrisome (ie, most suggestive of malignancy) should be reported and saved to the picture archiving and communication system.
Three types of time-enhancement curves are defined, according to whether initial enhancement (ie, enhancement within the first 2 minutes after contrast agent injection or until the curve changes direction) is slow, medium, or rapid. Delayed enhancement (ie, after the first 2 minutes or after the curve has changed direction) is characterized as persistent (type I), plateau (type II), or wash-out (type III) (18) (Fig 11).
View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11. Schematic drawing of time-enhancement curve types. The type I curve (dotted black line) indicates persistent enhancement throughout the examination. The curve for type II (solid black line) shows peak enhancement at 2–3 minutes after contrast agent injection, followed by a plateau. The type III curve (gray line) shows peak enhancement followed by washout with a steady decrease in signal intensity.
|
|
Rapid early enhancement, early washout, and peripheral rim enhancement are typical features of a malignancy. Weak enhancement, smooth mass margins, and continual contrast agent uptake are less suggestive of a malignancy. The likelihood of malignancy is 6% for lesions with type I curves, 64% for lesions with type II curves, and 87% for lesions with type III curves (18).
Generally, the most suspicious characteristics–either the morphologic features or the enhancement kinetics–will dictate the further management of a lesion.
For example, a lesion with benign morphologic features and a type III time-enhancement curve (ie, washout kinetics), or an enhanced lesion with malignant morphologic features regardless of the curve type, warrants biopsy.
Potential Pitfalls
Contrast Enhancement.—
Although enhancement is the best indicator of breast cancer (more than 90% of breast cancers appear strongly enhanced), it is nonspecific; there is a significant overlap in contrast enhancement characteristics between many benign and many malignant lesions. Benign lesions such as fibroadenoma, proliferative and nonproliferative fibrocystic change, inflammatory change, scar, sclerosing adenosis, lobular carcinoma in situ, and atypical ductal hyperplasia may appear strongly enhanced (29) (Fig 12). False-negative findings at breast MR imaging may occur in the presence of invasive lobular carcinoma, metastatic breast masses, low-grade intraductal carcinoma, and well-differentiated invasive ductal malignancies. Furthermore, a mucinous carcinoma may lack strong enhancement because the gelatinous tumor matrix may replace the characteristic solid enhancing component seen in other tumors.
View larger version (140K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12. Axial contrast-enhanced T1-weighted fat-suppressed image shows three enhanced, smooth, lobulated masses with low-signal-intensity internal septa. These findings represented fibroadenomas, which were confirmed at excisional biopsy.
|
|
Failure of Injection.—
Failure of the contrast agent injection (eg, infiltration of the contrast agent) may lead to the misinterpretation of breast MR images as negative. Before a negative result is reported, a contrast-enhanced appearance of the heart and internal mammary vessels on MR images should be verified to ensure that the contrast agent was properly administered.
Chemical Shift Artifact.—
The resonant frequencies of water and fat at 1.5 T differ by 3.5 ppm or 224 Hz, a difference that may cause misregistration between the fat image and the water image in the frequency-encoding direction. Depending on gradient strengths, the images may be misregistered by as much as several pixels. Chemical shift artifacts can be corrected by using fat suppression techniques or, if such techniques cannot be used, by increasing the bandwidth of image acquisition.
Flow-related Ghost Artifact.—
Ghost artifacts related to blood flow in the heart, aorta, and vena cava are recognizable by their high signal intensity and by their uniform spacing or blurring due to periodic motion. Ghost artifacts are propagated in the phase-encoding direction. The obscuration of breast tissue by such defects can be minimized by ensuring that the phase-encoding direction at breast MR imaging is not anterior-posterior.
Magnetic Susceptibility Artifact.—
Magnetic susceptibility artifacts may appear near metallic surgical clips or breast tissue marking clips on MR images. The artifacts are caused by degradation of the magnetic field and magnetic susceptibility uniformity in the vicinity of ferromagnetic materials (Fig 13).
Extramammary Findings
Although the breast MR imaging protocol is tailored to evaluate the breast parenchyma, breast MR images also may provide a gratuitous view of a portion of the upper abdomen and chest. On occasion, this unintended coverage may result in incidental hepatic, renal, osseous, pulmonary, lymphatic, or dermal findings that are clinically relevant.
Reporting the Results
Once the breast MR imaging study has been interpreted, it should be reported in accordance with the American College of Radiology practice guideline for communication of diagnostic radiology information (19). The terminology used in the report should be that standardized in the American College of Radiology breast MR imaging lexicon, and the report should include a final assessment code (28).
Conclusions
Contrast-enhanced breast MR imaging is a powerful tool in the breast imaging armamentarium. An awareness of proper imaging technique, potential pitfalls, and artifacts is critical to achieve accurate image interpretation.
When breast MR imaging is performed with the required technical factors and in the appropriate clinical setting, it is a highly sensitive and reasonably specific method for the detection of breast cancer.
The recently published proposed breast MR imaging indications and lexicon will help standardize the performance of breast MR imaging and image interpretation.
|
Acknowledgments
|
|---|
The corresponding author thanks Bradley N. Delman, MD, for technical support in the preparation of the manuscript.
|
Footnotes
|
|---|
Abbreviations: DCIS = ductal carcinoma in situ, ROI = region of interest, SNR = signal-to-noise ratio
SUPPLEMENTAL MATERIAL
A patient questionnaire to supplement this article is available online at radiographics.rsnajnls.org/cgi/content/full /26/5/1469/DC1.
|
References
|
|---|
- Ross RJ, Thompson JS, Kim K, Bailey RA. Nuclear magnetic resonance imaging and evaluation of human breast tissue: preliminary clinical trials. Radiology 1982;143:195–205.[Abstract/Free Full Text]
- El Yousef SJ, Alfidi RJ, Duchesneau RH, et al. Initial experience with nuclear magnetic resonance (NMR) imaging of the human breast. J Comput Assist Tomogr 1983;7:215–218.[Medline]
- McSweeney MB, Small WC, Cerny V, Sewell W, Powell RW, Goldstein JH. Magnetic resonance imaging in the diagnosis of breast disease: use of transverse relaxation times. Radiology 1984;153:741–744.[Abstract/Free Full Text]
- Merchant TE, Thelissen GR, de Graaf PW, Nieuwenhuizen CW, Kievit HC, Den Otter W. Application of a mixed imaging sequence for MR imaging characterization of human breast disease. Acta Radiol 1993;34(4):356–361.[Medline]
- Kuhl CK, Klaschik S, Mielcarek P, Gieseke J, Wardelmann E, Schild HH. Do T2-weighted pulse sequences help with the differential diagnosis of enhancing lesions in dynamic breast MRI? J Magn Reson Imaging 1999;9:187–196.[CrossRef][Medline]
- Jacobs MA, Barker PB, Bluemke DA, et al. Benign and malignant breast lesions: diagnosis with multiparametric MR imaging. Radiology 2003;229:225–232.[Abstract/Free Full Text]
- Heywang SH, Hahn D, Schmidt H, et al. MR imaging of the breast using Gd-DTPA. J Comput Assist Tomogr 1986;10:615–620.[Medline]
- Heywang SH, Bassermann R, Fenzl G, et al. MRI of the breast: histopathologic correlation. Eur J Radiol 1987;7:175–182.[Medline]
- Weinreb JC, Newstead G. MR imaging of the breast. Radiology 1995;196:593–610.[Abstract/Free Full Text]
- Davis PL, McCarty KS Jr. Sensitivity of enhanced MRI for detection of breast cancer: new, multicentric, residual, and recurrent. Eur Radiol 1997; 7(suppl 5):289–298.[CrossRef][Medline]
- Orel SG, Schnall MD. MR imaging of the breast for the detection, diagnosis, and staging of breast cancer. Radiology 2001;220:13–30.[Abstract/Free Full Text]
- Heywang-Kobrunner SH, Viehweg P, Heinig A, Kuchler C. Contrast-enhanced MRI of the breast: accuracy, value, controversies, solutions. Eur J Radiol 1997;24:94–108.[CrossRef][Medline]
- Heywang-Kobrunner SH, Dershaw DD, Schreer I. Diagnostic breast imaging. Stuttgart, Germany: Thieme, 2001.
- Morris EA. Breast cancer imaging with MRI. Radiol Clin North Am 2002;40:443–446.[CrossRef][Medline]
- Harms SE, Flamig DP, Hesley KL, et al. MR imaging of the breast with rotating delivery of excitation off resonance: clinical experience with pathologic correlation. Radiology 1993;187:493–501.[Abstract/Free Full Text]
- Harms SE. Staging for breast cancer treatment. In: Stark DD, Bradley WG, eds. Magnetic resonance imaging. 3rd ed. St Louis, Mo: Mosby, 1999; 321–333.
- Boetes C, Barentsz JO, Mus RD, et al. MR characterization of suspicious breast lesions with a gadolinium-enhanced TurboFLASH subtraction technique. Radiology 1994;193:777–781.[Abstract/Free Full Text]
- 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? Radiology 1999;211:101–110.[Abstract/Free Full Text]
- American College of Radiology. ACR practice guideline for the performance of magnetic resonance imaging (MRI) of the breast. In: Practice guideline and technical standards, 2005. Reston, Va: American College of Radiology, 2005; 341–346.
- Londero V, Bazzocchi M, Del Frate C, et al. Locally advanced breast cancer: comparison of mammography, sonography and MR imaging in evaluation of residual disease in women receiving neoadjuvant chemotherapy. Eur Radiol 2004;14:1371–1379.[Medline]
- Fischer U, Kopka L, Grabbe E. Breast carcinoma: effect of preoperative contrast-enhanced MR imaging on the therapeutic approach. Radiology 1999;213:881–888.[Abstract/Free Full Text]
- Liberman L, Morris EA, Kim CM, et al. MR imaging findings in the contralateral breast of women with recently diagnosed breast cancer. AJR Am J Roentgenol 2003;180:333–341.
Top
Abstract
LEARNING OBJECTIVES
