| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Breast Imaging Division, Box 3808, Duke University
Medical Center, Durham, NC 27710.
2 Department of Medicine, Oncology Division, Box 3893, Duke University Medical
Center, Durham, NC 27710.
3 Department of Pathology, Box 3712, Duke University Medical Center, Durham, NC
27710.
4 Department of Biostatistics and Bioinformatics, Box 3958, Duke University
Medical Center, Durham, NC 27710.
5 Department of Radiation Oncology, Radiation Physics Division, Box 3085, Duke
University Medical Center, Durham, NC 27710.
6 Department of Radiation Oncology, Box 3455, Duke University Medical Center,
Durham, NC 27710.
Received March 14, 2003;
accepted after revision May 21, 2003.
Supported by National Institutes of Health, National Cancer Institute grant
CA42745.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Twenty-one patients with locally advanced primary breast carcinoma underwent contrast-enhanced MRI before and after treatment with neoadjuvant anthracycline-based chemotherapy. For each patient, the maximum extent of the MRI abnormality was measured both before and after treatment. These measurements were subsequently compared with physical examination findings and histologic results to determine the ability of MRI to accurately reveal tumor extent after neoadjuvant chemotherapy.
RESULTS. MRI after chemotherapy showed a correlation coefficient of 0.75 with histology, which was better than physical examination (r = 0.61). MRI underestimated the extent of residual tumor in two patients by more than 1 cm (including one false-negative examination), was within 1 cm in 12 of 21 patients, and overestimated tumor extent by more than 1 cm in seven of 21 patients.
CONCLUSION. MRI can show residual malignancy after neoadjuvant chemotherapy better than physical examination, particularly in patients who have not had a complete clinical response to therapy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This study was undertaken to evaluate the ability of MRI to accurately show residual primary breast malignancy in women treated with neoadjuvant chemotherapy by comparing imaging data with histologic findings after surgery and to evaluate the limitations associated with the interpretation of breast MRI after neoadjuvant chemotherapy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The medical records of the 21 women were reviewed to determine the initial presentation of the breast carcinoma and the clinical response to chemotherapy.
MRI Technique
Each patient underwent contrast-enhanced MRI of the breast before and after
neoadjuvant chemotherapy administration. MRI was performed immediately before
the administration of the first cycle of chemotherapy, 1–2 weeks after
the completion of adjuvant chemotherapy, and within 2 weeks of planned
definitive surgery. All patients had previously undergone percutaneous core
needle breast biopsy 1–4 weeks before the initial MRI examination.
All MRI examinations were performed on a commercially available 1.5-T system (Signa, General Electric Medical Systems, Milwaukee, WI) using a dedicated breast coil (Liberty 5000, USA Instruments, Aurora, OH). The imaging sequences were similar for all patients and for the pre- and postchemotherapy examinations. After an axial localizer image was obtained, a sagittally oriented fat-suppressed fast spin-echo T2-weighted sequence was performed (TR/TE, 4,000/110; 512 x 512 matrix; 4.0-mm slice thickness; no gap). This sequence was followed by a T1-weighted fat-suppressed gradient-recalled echo sequence (TR/TE, variable; 512 x 512 matrix; 1- to 2-mm slice thickness; no gap). This sequence was carefully reviewed before contrast administration to ensure adequate and uniform fat saturation. Once this sequence was optimized, gadopentetate dimeglumine (Magnevist [0.1 mmol/L per kilogram of body weight], Berlex, Wayne, NJ) was administered IV by a rapid injector through an indwelling IV catheter. Contrast-enhanced imaging was initiated immediately after the contrast material was completely injected and consisted of three volume acquisitions through the breast. Two were performed in rapid succession immediately after contrast administration, and the third was performed after a 10-min delay from the initial injection. All contrast-enhanced imaging was performed with the same T1-weighted sequence optimized before contrast injection. Unenhanced and contrast-enhanced MRI examinations were performed in the sagittal plane, and image acquisition times ranged from 45 sec to 2 min per volume with an approximate total imaging time of 20 min per examination. Section thickness varied depending on the size of the breast and ranged from 1 to 2 mm without a gap, using a 512 x 256 matrix and a 16- to 18-cm field of view.
Prospective Breast MRI Interpretation
All MRI examinations were initially interpreted by one of four dedicated
breast imaging radiologists with breast MRI experience. Both hard-copy images
and soft-copy display were used in this interpretation. Images were evaluated
for areas of abnormal enhancement in the breast using previously described
interpretation criteria
[5–8].
In addition, associated findings of architectural distortion, skin thickening,
skin enhancement, nipple retraction, nipple enhancement, axillary adenopathy,
chest wall invasion, or pectoralis invasion were reported when shown.
Clinical, mammographic, laboratory, and histologic results of core needle
biopsy were available for all patients and were reviewed either before or
concurrent with the prospective MRI evaluation. MRI examinations after
chemotherapy were compared with the MRI examinations before chemotherapy to
assess interval change. All enhancing abnormalities that were indeterminate
for malignancy and that were separate from the known primary tumor were
regarded as suspicious. If breast conservation was being considered,
additional evaluation with either imaging-directed core needle biopsy or
MRI-guided localization followed by surgical biopsy was recommended.
Neoadjuvant Chemotherapy
Patients were treated preoperatively with one of the following regimens.
Most patients (n = 17) were treated with a combination of paclitaxel
(Taxol [100–175 mg/m2], Bristol-Myers Squibb, Princeton, NJ),
liposomal doxorubicin (Myocet [40–60 mg/m2], Elan
Pharmaceuticals, Princeton, NJ), and breast hyperthermia given over 3 hr,
every 21 days, for a total of four cycles. Four additional patients were
treated with the standard regimen of docetaxel (Taxotere [75
mg/m2], Aventis Pharmaceuticals, Bridgewater, NJ) and doxorubicin
given over 1 hr, every 21 days, for a total of four cycles. Informed consent
was obtained from all patients before chemotherapy administration.
Four hyperthermia treatments were given every 21 days as part of neoadjuvant local and systemic therapy. Hyperthermia treatment started immediately after injection of Myocet, with an overall hyperthermia dose goal to reach 41.0–41.5°C in greater than 90% of measured points for 60 min. Hyperthermia was administered to the tumor using microwaves and following standard techniques, which were individualized by anatomy and tumor location. Maximum allowable temperatures in adjacent normal tissues and tumor were 43°C and 48°C, respectively. During treatment, the applied power was adjusted as needed to reach but not exceed the desired temperatures, within patient tolerance. Thermal data were analyzed to determine the T90 (10th percentile of the temperature distribution) and T50 (median temperature) for each treatment and the T10 (90th percentile of the temperature distribution). These parameters were averaged over all treatments to obtain mean thermal data for each patient.
Data Collection and Analysis
For the purpose of this study, all MRIs were retrospectively reevaluated by
a single radiologist with extensive breast MRI experience. At the time of this
retrospective review, the radiologist was not provided with clinical
information, radiologic imaging, laboratory data, or pathology outcome. For
each patient, the studies performed before chemotherapy and then the studies
performed after chemotherapy studies were evaluated, first separately and then
concurrently. The image or images best depicting the abnormality were
selected, and the lesion was measured using electronic calipers. The maximum
dimension of the abnormally enhancing lesion or lesions corresponding to the
known malignancy was measured. This maximum dimension was chosen to represent
the extent of disease present in the breast. Thus, if the lesion consisted of
multiple adjacent abnormalities, the maximum dimension was not the sum of
their diameters, but a single measurement encompassing the lesions farthest
apart. The goal was to determine the extent of malignancy in the affected
breast. The distribution of abnormally enhancing lesions in the breast was
also characterized as focal (single discrete mass) or nonfocal (segmental,
regional, or diffuse) distribution. Interval change was assessed by directly
comparing similar images from studies performed before and after chemotherapy
and comparing measurements obtained from these similar images. In addition, a
0.3-cm region-of-interest cursor was positioned on the area of maximum signal
intensity within the malignancy on both the unenhanced and early
contrast-enhanced images and the percentage of peak enhancement was
calculated.
Statistical Analysis
Histologic measurement of tumor size was used as the gold standard and was
compared with tumor measurements from physical examination and from MRI both
before and after treatment. Pearson's correlation coefficients were calculated
to determine the association between the MRI and physical examination
measurements and histologic size. Linear regression analysis was performed to
further characterize the nature of the relationships between MRI, physical
examination, and histology. Paired Student's t test analysis was
performed to determine changes in lesion peak contrast enhancement before and
after treatment.
Histologic Evaluation of Lumpectomy and Mastectomy Specimens
Breast specimens were evaluated by routine pathologic examination. The
specimens were processed by serial gross sectioning at approximately 1-cm
intervals. The largest grossly identifiable tumor mass was measured, and
representative H and E–stained slides were examined to confirm the
presence of carcinoma. The closest margin was evaluated histologically, and
all axillary lymph nodes were completely submitted for microscopic
examination. Representative sections of grossly normal breast tissue were
examined from quadrants distant from the largest tumor mass. Any other
suspicious masses in the breast that were detected by gross examination were
also sampled for histologic examination.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
MRI accurately showed final tumor size to within 1 cm in 12 (57%) of 21 patients, underestimated tumor size by more than 1 cm in two patients (10%), and overestimated tumor size by more than 1 cm in seven patients (33%) (Figs. 3A, 3B, 4A, 4B, 5A, 5B). In the one false-negative MRI examination, a 0.5-cm invasive ductal carcinoma was not shown on MRI after treatment (Fig. 6A, 6B). In the one "false-positive" case, MRI showed a 4.3-cm region of patchy, mild enhancement after chemotherapy. Although no invasive carcinoma was shown in this case, the pathologist commented on multicentric in situ carcinoma present diffusely throughout the submitted blocks.
|
|
|
|
|
|
|
|
MRI lesion peak percentage of enhancement was measured before and after neoadjuvant therapy to assess changes in vascularity resulting from treatment. Paired Student's t test analysis shows a significant reduction in the percentage of peak enhancement within lesions after neoadjuvant therapy (p < 0.003). Before treatment, the mean percentage of peak enhancement was 113% (SD = 44%, range = 44–212%), whereas the mean percentage of peak enhancement after treatment was 56% (SD = 53%, range = 0–197%). On average, a 40% reduction in contrast enhancement was shown within lesions when comparing images obtained before and after neoadjuvant chemotherapy. No lesions showed an increase in percentage of peak enhancement after the administration of neoadjuvant chemotherapy.
Of the 21 patients, 17 were surgically treated with modified radical mastectomy and four underwent breast conservation surgery. Of the patients who opted for breast conservation surgery, three had resectable tumor with negative surgical margins after the initial surgery, and one had positive surgical margins that required further surgical intervention. None of the four women who underwent breast conservation surgery ultimately underwent mastectomy.
The distribution of abnormal enhancement changed in three of 21 patients between the pre- and posttreatment MRI examinations. One focal lesion was not present after treatment (true-negative study), one segmental abnormality before treatment appeared as a focal lesion after treatment, and one segmental abnormality seen before treatment was not visible after treatment (false-negative study). In the other patients, although the maximum lesion diameter had decreased substantially between the two MRI examinations, MRI after chemotherapy revealed measurable abnormalities without any change in the overall distribution.
Hyperthermia Treatment
The means of the T90, T50, and T10 temperatures for all patients were
39.5°C ± 0.8°C, 40.8°C ± 0.9°C, and 42.3°C
± 1.1°C, respectively. These temperatures were high enough to cause
changes in perfusion and vascular permeability that could lead to an increase
in liposomal uptake in tumor, but were not high enough to cause significant
tissue damage
[9–11].
Therefore, the MRI results should not have been affected by this treatment
aside from what would be expected by chemotherapy alone.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There are no universally accepted criteria for attempting breast conservation in patients with locally advanced breast cancer, although at least one study has shown that tumors exceeding 2 cm after chemotherapy are more likely to recur if breast conservation surgery is performed [12]. Others have suggested that the size and extent of the initial tumor are the most important parameters for determining surgical options after neoadjuvant chemotherapy. Regardless, it is clear that accurate tools for assessing tumor size and response to chemotherapy are important for optimizing the clinical management of patients receiving this increasingly used treatment regimen.
Several studies comparing clinical and imaging assessment of tumors treated with preoperative chemotherapy have shown that MRI is superior to clinical examination, mammography, and sonography for assessing tumor response to chemotherapy [13–16]. However, the gold standard for preoperative assessment of tumor response remains directed physical examination.
Contrast-enhanced MRI of the breast is highly sensitive for detecting invasive breast carcinoma [17, 18]. High-resolution fat-suppressed T1-weighted imaging performed before and immediately after gadolinium administration yields images with high spatial resolution and moderate temporal resolution that allow identification and characterization of abnormally enhancing lesions representing carcinoma. These MRI techniques have been shown capable of accurately depicting primary breast carcinoma and have become clinically useful for showing occult primary malignancy; multifocal, or multicentric, disease; and the extent of known primary carcinoma [17–22].
Although, in general, the existing literature supports MRI as a method for assessing tumor response, there is less agreement about whether MRI can accurately measure the size of residual tumor after chemotherapy. Two recent studies reported a very high correlation between histology and postchemotherapy MRI in the assessment of residual malignancy [15, 23]. A third study, however, reported a less favorable association and a substantial false-negative rate for breast MRI after chemotherapy administration [24]. In fact, in that study, Rieber et al. [24] observed a 33% false-negative rate for postchemotherapy MRI and attributed this result to the dramatic reduction or lack of contrast enhancement in the breast after treatment. In a more recent study, Rieber et al. [25] concluded that although MRI is excellent as a qualitative method for assessing response to neoadjuvant chemotherapy, it is unreliable for determining the size of residual tumor. These researchers found that MRI tended to underestimate the amount of tumor shown at histology and that this tendency is most prevalent among tumors that respond well to chemotherapy. Although Partridge et al. [23] observed a statistically significant reduction in peak enhancement on MRI after chemotherapy, this finding did not result in significant underestimation of residual malignancy. In fact, according to their analysis, breast MRI after chemotherapy tended to overestimate the extent of residual malignancy.
We evaluated 21 patients who had locally advanced breast cancer with MRI both before and after the completion of neoadjuvant chemotherapy to determine how MRI after treatment compared with physical examination and histology in the assessment of disease extent. For the purposes of our study, a combination of the gross and microscopic histologic evaluations of the surgically excised breast specimen was used as the gold standard. We compared the largest tumor dimension shown by MRI with the size shown at physical examination and histology.
Like the data reported in previous studies [16, 17], our data show a statistically significant reduction in peak contrast enhancement after neoadjuvant chemotherapy. Although we used an MRI protocol similar to that described by Partridge et al. [23], we encountered several cases in which MRI underestimated residual malignancy, one of which resulted in a false-negative study. On the basis of this experience, we believe that the decreased contrast enhancement observed on postchemotherapy MRI can result in disease underestimation. Hayes et al. [26] showed that many primary breast carcinomas exhibit reduced contrast enhancement after chemotherapy. They hypothesized that this phenomenon may be caused directly by an antivascular effect of the chemotherapy agent or indirectly by tumor cell death, which may decrease local concentration of vascular growth factors. Our data concur with prior studies that observed significant decreased peak enhancement within breast tumors after chemotherapy, and our data also support the hypothesis that this decreased enhancement can result in the underestimation of residual malignancy on MRI, even a false-negative MRI.
In our series, however, there were more cases in which MRI overestimated the size of residual malignancy compared with histology. We overestimated tumor extent on the basis of postchemotherapy MRI in 11 (52%) of 21 cases; in seven of these cases, tumor extent was overestimated 1 cm or more. Five of these cases were in patients who presented with extensive, nonfocal abnormalities on the initial MRI examination. These results are similar to the findings of Partridge et al. [23], who reported an overall tendency for MRI to overestimate disease extent. Despite the statistically significant decrease in lesion enhancement on MRI after neoadjuvant chemotherapy, overestimation—not underestimation—of tumor size on MRI was more common in our series.
Our data suggest that MRI more closely depicts the size of malignancy shown at histologic evaluation than does physical examination. In this study, however, the difference between the correlation coefficients for MRI and histology and for physical examination and histology was substantially less than that reported by either Partridge et al. [23] or Weatherall et al. [15], and this difference was not statistically significant. Partridge et al. evaluated 52 lesions on MRI performed before and after neoadjuvant chemotherapy and showed a 0.89 correlation between posttreatment MRI and histology [23]. Similarly, in their evaluation of 20 lesions, Weatherall et al. [15] reported a 0.93 correlation coefficient between MRI and histology. Our data showed not only a lower correlation (0.75) between posttreatment MRI and histology, but also a smaller difference between MRI and physical examination (0.75 vs 0.65, respectively). Several possible explanations for this discrepancy between our study and others are the small sample size of our study, the large initial tumor sizes, the high percentage of nonfocal lesions, differences in histologic tumor types, and differences used to determine tumor extent on MRI.
In our series, 15 (71%) of 21 lesions were segmental, regional, or diffuse, comprising a higher percentage of nonfocal lesions than the series reported by Partridge et al. [23]. Our data suggest that nonfocal lesions are more difficult to assess than focal lesions on MRI after neoadjuvant chemotherapy. The focal lesions in our series (6/21) showed a 0.85 correlation coefficient with histology, whereas the nonfocal lesions showed a 0.69 correlation coefficient. In addition, MRI was much better than physical examination for assessing focal lesions, but only marginally better when the lesion was nonfocally distributed. Five of seven cases in which MRI overestimated disease extent by more than 1 cm were nonfocal lesions. One problem with the MRI evaluation of nonfocal lesions is that they can manifest as small areas of patchy enhancement, and this appearance can overlap with that of nonmalignant areas of enhancement, making distinction between tumor and normal tissue difficult. In our series, MRI evaluation of nonfocal malignancies correlated less well with histology than MRI evaluation of lesions that were focal.
Another explanation for the relatively poor correlation observed between MRI and histology among nonfocal lesions is that standard histologic evaluation may underestimate the true extent of malignancy (Fig. 7A, 7B). In fact, we believe this possibility is a major limitation of this and other studies attempting to assess the accuracy of MRI by comparing MRI with histology. The routine evaluation of mastectomy specimens provides only limited opportunity for detailed correlation with MRI studies, especially in the post–neoadjuvant chemotherapy setting. Patients with good clinical responses often have innumerable small foci of residual invasive carcinoma spread through the original tumor bed. By standard staging criteria, tumor size is reported as the largest single focus of invasive carcinoma, which may be very small, even when the tumor bed (total field of tumor) is very large. In addition, the gross examination in the neoadjuvant setting is often misleading—large mass lesions can be histologically sterile, whereas grossly normal breast may have extensive residual invasive or in situ carcinoma. Measurement of intraductal tumor size, although commonly performed in excisional biopsies, is rarely done in mastectomies because the number of microscopic sections needed is prohibitive and the information gained would not affect clinical management. Finally, small lesions seen on MRI may be missed on gross examination because of the relatively large (1 cm) sectioning interval, especially if the lesion is not easily palpable. Thus, there are substantial impediments to correlating many of the MRI findings with the results of routine pathologic evaluation of breast specimens, especially in the post–neoadjuvant chemotherapy setting.
|
|
In conclusion, in our study, MRI showed residual malignancy slightly better than physical examination after neoadjuvant chemotherapy. However, the difference between these two modalities was smaller than previously reported and did not reach statistical significance. Our data analysis suggests that MRI may be particularly useful in assessing residual malignancy when physical examination suggests minimal or partial clinical response to chemotherapy or when physical examination is difficult to assess. MRI performed after neoadjuvant chemotherapy is, in general, able to show the extent of residual malignancy; however, both false-negative and false-positive findings are possible. Therefore, we believe that caution should be used in basing surgical treatment options solely on the results of this examination. Moreover, future studies should strongly consider improved strategies for pathologic confirmation of MRI-identified breast lesions.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Elsamaloty, M. S. Elzawawi, S. Mohammad, and N. Herial Increasing Accuracy of Detection of Breast Cancer with 3-T MRI Am. J. Roentgenol., April 1, 2009; 192(4): 1142 - 1148. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. A. Buchholz, C. D. Lehman, J. R. Harris, B. A. Pockaj, N. Khouri, N. F. Hylton, M. J. Miller, T. Whelan, L. J. Pierce, L. J. Esserman, et al. Statement of the Science Concerning Locoregional Treatments After Preoperative Chemotherapy for Breast Cancer: A National Cancer Institute Conference J. Clin. Oncol., February 10, 2008; 26(5): 791 - 797. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. K. Kuhl Current Status of Breast MR Imaging * Part 2. Clinical Applications Radiology, September 1, 2007; 244(3): 672 - 691. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Loibl, G. von Minckwitz, G. Raab, J.-U. Blohmer, S. D. Costa, B. Gerber, H. Eidtmann, S. Petrich, J. Hilfrich, C. Jackisch, et al. Surgical Procedures After Neoadjuvant Chemotherapy in Operable Breast Cancer: Results of the GEPARDUO Trial Ann. Surg. Oncol., November 1, 2006; 13(11): 1434 - 1442. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Tozaki, T. Kobayashi, S. Uno, K. Aiba, H. Takeyama, H. Shioya, I. Tabei, Y. Toriumi, M. Suzuki, and K. Fukuda Breast-Conserving Surgery After Chemotherapy: Value of MDCT for Determining Tumor Distribution and Shrinkage Pattern Am. J. Roentgenol., February 1, 2006; 186(2): 431 - 439. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Shah, S. K. Shah, and K. V. Greatrex Current Role of Magnetic Resonance Imaging in Breast Imaging: A Primer for the Primary Care Physician J Am Board Fam Med, November 1, 2005; 18(6): 478 - 490. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Roubidoux, G. L. LeCarpentier, J. B. Fowlkes, B. Bartz, D. Pai, S. P. Gordon, A. F. Schott, T. D. Johnson, and P. L. Carson Sonographic Evaluation of Early-Stage Breast Cancers That Undergo Neoadjuvant Chemotherapy J. Ultrasound Med., July 1, 2005; 24(7): 885 - 895. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. C. Partridge, J. E. Gibbs, Y. Lu, L. J. Esserman, D. Tripathy, D. S. Wolverton, H. S. Rugo, E. S. Hwang, C. A. Ewing, and N. M. Hylton MRI Measurements of Breast Tumor Volume Predict Response to Neoadjuvant Chemotherapy and Recurrence-Free Survival Am. J. Roentgenol., June 1, 2005; 184(6): 1774 - 1781. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Koomen, E. D. Pisano, C. Kuzmiak, D. Pavic, and R. McLelland Future Directions in Breast Imaging J. Clin. Oncol., March 10, 2005; 23(8): 1674 - 1677. [Full Text] [PDF]
|
|
|
|
|
|
L. A. Newman and H. M. Kuerer Advances in Breast Conservation Therapy J. Clin. Oncol., March 10, 2005; 23(8): 1685 - 1697. [Full Text] [PDF]
|
|
|
|
 |
|
 N. E. Davidson and M. Morrow Sometimes a Great Notion--An Assessment of Neoadjuvant Systemic Therapy for Breast Cancer J Natl Cancer Inst, February 2, 2005; 97(3): 159 - 161. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= MRI) and best-fit line between MRI, physical
examination, and histology. Solid line = identity, dashed line = MRI, dotted
line = physical examination.
