A physics-based coordinate transformation for 3-D image matching.

Many image matching schemes are based on mapping coordinate locations, such as the locations of landmarks, in one image to corresponding locations in a second image. A new approach to this mapping (coordinate transformation), called the elastic body spline (EBS), is described. The spline is based on a physical model of a homogeneous, isotropic three-dimensional (3-D) elastic body. The model can approximate the way that some physical objects deform. The EBS as well as the affine transformation, the thin plate spline [1], [2] and the volume spline [3] are used to match 3-D magnetic resonance images (MRI's) of the breast that are used in the diagnosis and evaluation of breast cancer. These coordinate transformations are evaluated with different types of deformations and different numbers of corresponding (paired) coordinate locations. In all but one of the cases considered, using the EBS yields more similar images than the other methods.

Detection with MR imaging of residual tumor in the breast soon after surgery.

OBJECTIVE: Difficulties in the preoperative assessment of tumor size and extent result in a positive pathologic margin in up to 70% of patients undergoing breast conservation surgery. Although positive margins usually require reexcision, the location and extent of surgery required are often difficult to establish by current imaging techniques. We investigated the accuracy of three-dimensional rotating delivery of excitation off resonance (3D RODEO) MR imaging of the breast in revealing the presence and extent of residual tumor within the breast soon after surgery. MATERIALS AND METHODS: Nineteen patients who had undergone lumpectomy or excisional biopsy were evaluated with contrast-enhanced 3D RODEO MR imaging of the breast within 10 months after surgery. The MR imaging results were correlated with serial-sectioned mastectomy or partial mastectomy specimens from 18 patients and with a clinical and mammographic follow-up examination in one patient. RESULTS: We found that 3D RODEO MR imaging accurately revealed the presence or absence and the location and extent of recurrent tumor in 15 of the 18 patients who had pathologic confirmation. Of the three MR imaging-pathology mismatches, two had irregular or nodular enhancement that corresponded to microabscesses. The third mismatch showed multicentric disease on MR imaging but only single-quadrant lobular carcinoma at pathologic examination. Our 19th patient showed no evidence of recurrent tumor on MR imaging or at 2-year follow-up clinical and mammographic examinations. CONCLUSION: MR imaging with 3D RODEO technique correctly revealed the presence or absence, the location, and the extent of recurrent tumor in 84% of patients who had recently undergone breast surgery.

MR imaging in the management before surgery of lobular carcinoma of the breast: correlation with pathology.

OBJECTIVE: Our objective was to investigate the use of MR imaging in preoperative staging and characterization of lobular carcinoma. MATERIALS AND METHODS: MR imaging studies and mammographic studies in 20 patients with infiltrating lobular carcinoma were evaluated and correlated with pathology findings on serially sectioned tissue. The MR images and mammograms were reviewed retrospectively by three independent examiners unaware of the clinical, imaging, and pathology findings. RESULTS: The extent of disease found pathologically correlated with that predicted by MR imaging studies in 85% of patients, compared with a 32% correlation (p < .0001) with mammographic studies. Interobserver agreement on lesion morphology and extent of disease was higher for MR imaging (91% and 100%, respectively) than for mammography (64% and 91%, respectively). The retrospective MR readings did not differ from the prospective reports. No false-positive lymphadenopathy was interpreted on MR imaging. Lymph nodes having metastatic lobular carcinoma on the pathology examination were missed on MR imaging in four patients. CONCLUSION: MR imaging is significantly more accurate than mammography in determining the extent of disease and characterizing the morphology of infiltrating lobular carcinoma. MR imaging may play a role in preoperative planning, especially when breast conservation is being considered.

Three-dimensional RODEO breast MR imaging of lesions containing ductal carcinoma in situ.

PURPOSE: To assess whether rotating delivery of excitation off resonance (RODEO) breast magnetic resonance (MR) imaging can help detect ductal carcinoma in situ (DCIS) lesions, determine tumor extent, and differentiate pure DCIS from DCIS with an invasive component. MATERIALS AND METHODS: Twenty-two patients with DCIS lesions were evaluated with three-dimensional RODEO MR imaging. Nineteen patients had available mammograms for review. RESULTS: MR imaging enabled detection of all 22 cases of DCIS, DCIS with microinvasion, or invasive ductal carcinoma with extensive intraductal component. A clumped enhancement pattern was seen on MR images in all cases of pure DCIS. Spiculated enhancement was seen in four of six (67%) patients who had DCIS with microinvasion and in nine of 11 (82%) who had invasive ductal carcinoma with extensive intraductal component. RODEO MR imaging enabled accurate determination of tumor extent in 21 of 22 (95%) patients. Mammography depicted 18 of 19 DCIS lesions. No mammographic feature helped differentiate pure DCIS from DCIS with microinvasion. Mammography enabled accurate determination of tumor extent in 14 of 19 (74%) patients. CONCLUSION: Three-dimensional RODEO MR imaging can be an adjunct to mammography because of its ability to enable better determination of tumor extent and differentiation of pure DCIS from DCIS with an invasive component.

Silicone-suppressed 3D MRI of the breast using rotating delivery of off-resonance excitation.

OBJECTIVE: A new silicone-suppressed MR technique was developed, and its efficacy in identifying free silicone and differentiating it from other breast tissues was investigated. MATERIALS AND METHODS: Silicone-suppressed MRI was performed using the RODEO (rotating delivery of excitation off-resonance) pulse sequence, which selectively eliminated signal from the narrow range of (CH3)4Si resonance. Ninety breasts in 61 patients were evaluated with both a fat-suppressed 3D MR sequence and a silicone-suppressed 3D MR sequence. RESULTS: Extracapsular free silicone and silicone injections demonstrated a unique appearance compared with normal breast tissue in all cases. Magnetic resonance identified free silicone in 26 breasts, 10 of which were confirmed pathologically or from a history of previous silicone injections. No free silicone was present on MR in 64 breasts; 8 of these were confirmed by biopsy or mastectomy as showing no evidence of free silicone. This technique was useful in evaluation of prosthesis integrity, free silicone, focal palpable or mammographic lesions, and the breast with silicone injections. CONCLUSION: Silicone-suppressed RODEO MRI of the breast can accurately identify free and intracapsular silicone and can distinguish silicone from other tissues. This provides unique information about the breast in a number of specific applications.

MR imaging of the breast: current status and future potential.

Current and potential roles for MR imaging in the management of breast disorders are reviewed along with the specific technical requirements for each application. Major topics include (1) evaluating breasts before biopsy to reduce the number of surgical biopsies for benign lesions, (2) staging of breast carcinoma in breast conservation candidates, (3) evaluating breasts with inconclusive conventional imaging, (4) coordination of minimally invasive surgery, and (5) evaluating silicone implant integrity.

Staging of breast cancer with MR imaging.

The vastly improved sensitivity of new MR methods can be used to define disease within the breast that cannot be seen with conventional breast imaging methods. Breast MR imaging is expected to have a significant role for the staging of breast cancer in breast conservation candidates. Ultimately, breast MR will be integrated with interstitial laser photocoagulation as a treatment method for breast cancer.

MR imaging of the breast with rotating delivery of excitation off resonance: clinical experience with pathologic correlation.

An investigative study was undertaken to determine the potential for a new magnetic resonance (MR) imaging technique, RODEO (rotating delivery of excitation off resonance), for use as a diagnostic imaging tool for the breast. The RODEO technique provides fat suppression with T1 weighting and is ideal for gadolinium-enhanced breast imaging. It is a short repetition time, steady-state sequence for high-resolution three-dimensional acquisitions and provides a clinically efficient imaging time of approximately 5 minutes for 128 sections. Imaging findings were correlated with serially sectioned pathologic specimens in 30 breasts with 47 malignant and 27 benign lesions. MR imaging had a sensitivity of 94% and a specificity of 37%. MR imaging depicted additional cancers not seen at mammography in 11 of the 30 patients (37%). The lesions not seen at mammography varied in size from 3 mm to 12 cm. RODEO MR imaging may be used to improve diagnosis of breast cancer in patients with mammographically dense breasts or silicone implants/injections and to stage disease in patients who are candidates for lumpectomy.

Fat-suppressed three-dimensional MR imaging of the breast.

Rotating delivery of excitation off-resonance (RODEO) is a new magnetic resonance (MR) imaging pulse sequence that uses a jump return sine excitation on fat resonance to produce fat-suppressed, T1-weighted images. New three-dimensional MR imaging techniques were used to examine 57 women with abnormalities suspicious for breast cancer. MR imaging findings were compared with those of mammography in all cases and with those of other imaging techniques when appropriate. Thirty-five specimens obtained at mastectomy were analyzed with rigorous pathologic examination that included imaging of the entire breast at 5-mm incremental sections. Histologic confirmation was obtained in 76 lesions in 47 patients. MR imaging helped detect 100% of malignant lesions, whereas mammography produced 33% false-negative findings. The use of RODEO in breast imaging is in the early investigational phases, but it has potential for supplementing mammography in the diagnosis of breast cancer.

MR imaging of the breast.

The current strategy for breast cancer treatment involves early detection of the neoplasm before it has metastasized outside the breast, and surgical treatment of the lesion that minimizes deformity. Conventional methods of diagnostic imaging of the breast, including mammography, sonography, and galactography, do not adequately address clinical needs with regard to lesion characterization and staging. Magnetic resonance (MR) imaging has been proposed as a modality that may address these needs. The potential clinical uses of MR imaging are defined in terms of clinical needs and technologic requirements. Applications of MR imaging in detection of breast lesions can be divided into two major approaches based on the clinical questions to be answered: (a) improved specificity and (b) improved sensitivity. These approaches are defined by specific sets of clinical needs and have substantially different technologic requirements. The specificity approach is used to reduce the number of biopsies performed to confirm false-positive mammographic findings. The MR imaging device that is used to improve specificity must visualize only the lesions that are seen mammographically. The sensitivity approach requires the visualization of lesions not identified at mammography, so that breast disease can be more effectively staged for treatment. The technologic requirements for the sensitivity approach are considerably more rigorous because all lesions must be identified. The advantages and disadvantages of each of these approaches and the potential clinical ramifications are described.

Magnetic resonance angiography. Application to the peripheral circulation.

Magnetic resonance angiography (MRA) for application in the musculoskeletal system is progressing at a rapid pace. The clinical role of this new technique is still being developed, although a number of uses have already become well recognized, and MRA is diffusing into clinical practice. This article describes the basic fundamentals of MRA as it applies to the peripheral circulation and to the evaluation of potential clinical procedures.

Magnetic resonance imaging of the breast.

The diagnosis and treatment of breast cancer is dependent upon accurate depiction of the disease by diagnostic imaging. In a number of clinical situations, conventional breast imaging does not adequately address these diagnostic needs. New magnetic resonance imaging (MRI) methods developed specifically for breast diagnosis may provide the additional capability needed to fill the gap between clinical needs and the information obtained by conventional breast imaging methods. Fat-suppressed 3D MRI has demonstrated improved sensitivity over routine breast imaging methods. MRI can also be used to differentiate between certain benign but mammographically suspicious lesions and cancer. The potential clinical roles of MRI are reviewed with clinical examples. Pitfalls in the use of MRI are defined. The problems encountered with the implementation of MRI in a clinical setting are outlined and future advances predicted.

Magnetization transfer contrast in fat-suppressed steady-state three-dimensional MR images.

We demonstrate that magnetization transfer contrast can be used to improve the diagnostic utility of fat-suppressed steady-state three-dimensional gradient-recalled images. Fat suppression is achieved using a "jump-return" pair of contiguous shaped pulses. No time interval exists between the pulses, and no RF echo is generated. The sequence normally produces images with "density" weighting. Preparation of the spin magnetization with off-resonance frequency-selective excitation creates magnetization transfer contrast which attenuates signal intensity in proportion to the exchange rate of magnetization from free water with magnetization from water bound to macromolecules or protons that have restricted mobility. The resulting images have excellent fat suppression with low sensitivity to motion since no subtraction is used. In addition, the mechanism of signal attenuation is independent of paramagnetic effects, and addition of Gd-DTPA produces signal enhancement from vascularized regions of tissue. Examples are presented for the knee and breast, where the observation of pathology with signal enhancement from Gd-DTPA is improved over conventional 3D fat-suppressed images.

Three-dimensional gadolinium-enhanced MR imaging of the breast: pulse sequence with fat suppression and magnetization transfer contrast. Work in progress.

A pulse sequence with magnetization transfer contrast and fat suppression was used in three-dimensional magnetic resonance imaging of the breast. Two healthy volunteers, one person with silicone implants, and 12 patients with clinical and/or mammographic findings suspicious for malignancy were evaluated prior to and following infusion of gadopentetate dimeglumine. Imaging time was approximately 7 minutes for each set of data (128 sections). Final voxel dimensions ranged from 1.4 x 0.8 x 0.8 mm to 1.6 x 0.9 x 0.9 mm. All carcinomas, including ductal and lobular types, were enhanced before and after infusion of contrast medium. Multifocal carcinoma and inflammatory carcinoma could be clearly visualized. Enhancement was not evident in patients with fat necrosis (n = 1) or scar (n = 1). Fibrocystic changes in one patient were visible as areas of increased signal intensity on preinfusion images. Resolution and contrast of MR images obtained with this pulse sequence appeared to be improved over that achieved with conventional breast MR imaging techniques. This method has the potential to supplement conventional diagnostic methods in the evaluation of breast disease.

In vivo proton magnetic resonance spectroscopy studies of human brain.

In vivo localized proton magnetic resonance spectroscopy (MRS) studies of brain were performed on eighteen normal subjects using the stimulated echo (STE) sequence. The absolute concentrations and proton relaxation times of N-acetyl aspartate (NAA), total creatine (Cr) and choline (Cho) were estimated. The MRS data was quantitatively analyzed for repeatability and intersubject variability. Quantitative analysis indicates excellent spectral repeatability. Significant intersubject variations in [NAA] and [Cr] have been observed while the intersubject variability in [Cho] has been found to be fairly small. Significant intensity distortions have been observed for mixing times longer than 50 msec.

One-dimensional spectroscopic imaging with stimulated echoes: phantom and human leg studies.

One-dimensional phase encoding was incorporated in the stimulated echo single voxel localization sequence for in vivo proton spectroscopic studies. Phantom studies were performed to assess the effect of the number of phase encoding steps on the spectral contamination from the adjacent volumes. Both water suppressed and unsuppressed spectra were obtained in reasonable acquisition times from various regions in the human leg with a spatial resolution of around 1.1 cm3.