Real-time fast strain-encoded magnetic resonance imaging to evaluate regional myocardial function at 3.0 Tesla: comparison to conventional tagging

PURPOSE: To compare the utility of the real-time technique fast strain-encoded magnetic resonance imaging (fast-SENC) for the quantification of regional myocardial function to conventional tagged magnetic resonance imaging (MRI). MATERIALS AND METHODS: Healthy volunteers (N = 12) and patients with heart failure (N = 7) were examined using tagged MRI and fast-SENC at 3.0T. Circumferential strain was measured using fast-SENC in six endo- and six subepicardial regions in the basal-, mid-, and apical-septum and the basal-, mid-, and apical-lateral wall from the four-chamber view. These measurements were plotted to tagging, in corresponding myocardial segments. RESULTS: Peak systolic strain (Ecc) and early diastolic strain rate (Ecc/second) acquired by fast-SENC correlated closely to tagged MRI (r = 0.90 for Ecc and r = 0.91 for Ecc/second, P < 0.001 for both). Both fast-SENC and tagging identified differences in regional systolic and diastolic function between normal myocardium and dysfunctional segments in patients with heart failure (for fast-SENC: Ecc = -21.7 +/- 2.7 in healthy volunteers vs. -12.8 +/- 4.2 in hypokinetic vs. 0.6 +/- 3.8 in akinetic/dyskinetic segments, P < 0.001 between all; Ecc/second = 104 +/- 20/second in healthy volunteers vs. 37 +/- 9/second in hypokinetic vs. -16 +/- 15/second in akinetic/dyskinetic segments, P < 0.001 between all). Quantitative analysis was more time-consuming for conventional tagging than for fast-SENC (time-spent of 3.8 +/- 0.7 minutes vs. 9.5 +/- 0.7 minutes per patient, P < 0.001). CONCLUSION: Fast-SENC allows the rapid and accurate quantification of regional myocardial function. The information derived from fast-SENC during a single heartbeat seems to be superior or equal to that acquired by conventional tagging during several heart cycles and prolonged breathholds.     

Coronary artery wall imaging: initial experience at 3 Tesla

Purpose

To assess the feasibility of black‐blood turbo spin‐echo imaging of the left anterior descending coronary artery wall at 3 Tesla under free‐breathing and breath‐hold conditions.

Materials and Methods

Proton density‐weighted black‐blood turbo spin‐echo imaging of the left anterior descending coronary artery was performed on 15 volunteers on a 3 T whole body scanner with an eight channel phased array coil. Volunteers were imaged during free‐breathing (with navigators, N = 5), or with breath‐hold (N = 5), or both (N = 2). Imaging was not possible in three volunteers due to either gradient or radiofrequency (RF) coupling with the electrocardiogram (ECG). Images were analyzed to determine coronary artery wall thickness, wall area, lumen diameter, and lumen area. Signal‐to‐noise and contrast‐to‐noise ratios were calculated.

Results

Coronary artery wall thickness, wall area, lumen diameter, and lumen area measurements were consistent with previous magnetic resonance (MR) measurements of the coronary wall at 1.5 Tesla.

Conclusion

Coronary wall imaging using free‐breathing and breath‐hold two‐dimensional black‐blood TSE is feasible at 3 T. Further improvement in resolution and image quality is required to detect and characterize coronary plaque. J. Magn. Reson. Imaging 2005;21:128–132. © 2005 Wiley‐Liss, Inc.

     

Strain‐encoded (SENC) magnetic resonance imaging to evaluate regional heterogeneity of myocardial strain in healthy volunteers: Comparison with conventional tagging

Purpose

To evaluate the ability of strain‐encoded (SENC) magnetic resonance imaging (MRI) for regional systolic and diastolic strain analysis of the myocardium in healthy volunteers.

Materials and Methods

Circumferential and longitudinal peak systolic strain values of 75 healthy volunteers (35 women and 40 men, mean age 44 ± 12 years) were measured using SENC at 1.5T. MR tagging was used as the reference standard for measuring regional function. Diastolic function was assessed in the 10 youngest (24 ± 8 years) and 10 oldest (62 ± 5 years) subjects.

Results

Peak strain values assessed with SENC were comparable to those obtained by MR tagging, showing narrow limits of agreement (limits of agreement ?5.6% to 8.1%). Regional heterogeneity was observed between different segments of the left ventricle (LV) by both techniques (P < 0.001). Longitudinal strain obtained by SENC was also heterogenous (P < 0.001). Interestingly, no age‐ or gender‐specific differences in peak systolic strain were observed, whereas the peak rate of relaxation of circumferential strain rate was decreased in the older group.

Conclusion

SENC is a reliable tool for accurate and objective quantification of regional myocardial systolic as well as diastolic function. In agreement with tagged MRI, SENC detected slightly heterogeneous myocardial strain within LV segments. J. Magn. Reson. Imaging 2009;29:99–105. © 2008 Wiley‐Liss, Inc.

     

Myocardial function in infarcted and remote regions early after infarction in man: Assessment by magnetic resonance tagging and strain analysis

Early after infarction in the perfusion bed of the left anterior descending coronary artery, cine MRI with spatial modulation of magnetization (SPAMM) tagging (7-mm grid) was used for short- and long-axis cardiac imaging. Two-dimensional strain analysis of triangular finite elements was performed between end-diastole and end-systole. Patients (n = 10) were compared with age-matched healthy subjects (n = 8). The anteroseptal region at midventricular level was considered representative for “infarcted” and the posterolateral region at basal level was considered “remote”. The left ventricular end-diastolic volume index was larger in the patients (69 ± 15 ml/m² versus 56 ± 4 ml/m², P < 0.05). Short-axis images showed in the infarcted region a decrease of first principal strain (greatest systolic lengthening: 1.10 ±. 06 versus 1.27 ± 0.04, P < 0.0001), and in the remote region an increase (1.48 ± 0.11 versus 1.36 ± 0.07, P < 0.025). The lateral and inferior ventricular regions at mid- and basal levels were found to function normally. Long-axis images yielded similar results. Early after infarction, regions with dysfunction, normal function, and hyperfunction can be delineated with MR tagging. The compensatory increased contraction in the remote region is possibly triggered by the Frank-Starling mechanism.     

Cardiac magnetic resonance parallel imaging at 3.0 Tesla: technical feasibility and advantages

PURPOSE: To quantify changes in signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), specific absorption rate (SAR), RF power deposition, and imaging time in cardiac magnetic resonance imaging with and without the application of parallel imaging at 1.5 T and 3.0 T. MATERIALS AND METHODS: Phantom and volunteer data were acquired at 1.5 T and 3.0 T with and without parallel imaging. RESULTS: Doubling field strength increased phantom SNR by a factor of 1.83. In volunteer data, SNR and CNR values increased by factors of 1.86 and 1.35, respectively. Parallel imaging (reduction factor = 2) decreased phantom SNR by a factor of 1.84 and 2.07 when compared to the full acquisition at 1.5 T and 3.0 T, respectively. In volunteers, SNR and CNR decreased by factors of 2.65 and 2.05 at 1.5 T and 1.99 and 1.75 at 3.0 T, respectively. Doubling the field strength produces a nine-fold increase in SAR (0.0751 to 0.674 W/kg). Parallel imaging reduced the total RF power deposition by a factor of two at both field strengths. CONCLUSIONS: Parallel imaging decreases total scan time at the expense of SNR and CNR. These losses are compensated at higher field strengths. Parallel imaging is effective at reducing total power deposition by reducing total scan time.     

Myocardial perfusion and function in dogs with moderate coronary stenosis

MRI studies of first-pass contrast enhancement with polylysine-Gd-DTPA and myocardial tagging using spatial modulation of magnetization (SPAMM) were performed to assess the feasibility of a combined regional myocardial blood flow and 2D deformation exam. Instrumented closed-chest dogs were imaged at a baseline control state (Cntl) followed by two interventions: moderate coronary stenosis (St) achieved by partial occlusion of the left anterior descending (LAD) and moderate coronary stenosis with dobutamine loading (StD). Hypoperfusion of the anterior region (ANT) of the myocardium (LAD distribution) relative to the posterior wall (POS) based on the upslope of the signal intensity time curve from the contrast-enhanced MR images was demonstrated only with dobutamine loading (ANT:POS Cntl=1.077 ± 0.15 versus ANT:POS StD=0.477 ± 0.11, P<0.03) and was confirmed with radio-labeled microspheres measurements (ANT:POS Cntl=1.18 ± 0.2 ml/min/g versus ANT:POS StD=0.44 ± 0.1 ml/min/g; P<0.002). Significant changes in regional myocardial shortening were only seen in the StD state (P<0.02); the anterior region showed impaired myocardial shortening with dobutamine loading (P=NS), whereas the nonaffected POS region showed a marked increase in shortening when compared with Cntl (Cntl=0.964 ± 0.02 versus StD=0.884 ± 0.03; P<0.001). These results demonstrate that an integrated quantitative assessment of regional myocardial function and semiquantitative assessment of myocardial blood flow can be performed noninvasively with ultrafast MRI.     

Multislice dark-blood carotid artery wall imaging: a 1.5 T and 3.0 T comparison

PURPOSE: To compare two multislice turbo spin-echo (TSE) carotid artery wall imaging techniques at 1.5 T and 3.0 T, and to investigate the feasibility of higher spatial resolution carotid artery wall imaging at 3.0 T. MATERIALS AND METHODS: Multislice proton density-weighted (PDW), T2-weighted (T2W), and T1-weighted (T1W) inflow/outflow saturation band (IOSB) and rapid extended coverage double inversion-recovery (REX-DIR) TSE carotid artery wall imaging was performed on six healthy volunteers at 1.5 T and 3.0 T using time-, coverage-, and spatial resolution-matched (0.47 x 0.47 x 3 mm3) imaging protocols. To investigate whether improved signal-to-noise ratio (SNR) at 3.0 T could allow for improved spatial resolution, higher spatial resolution imaging (0.31 x 0.31 x 3 mm3) was performed at 3.0 T. Carotid artery wall SNR, carotid lumen SNR, and wall-lumen contrast-to-noise ratio (CNR) were measured. RESULTS: Signal gain at 3.0 T relative to 1.5 T was observed for carotid artery wall SNR (223%) and wall-lumen CNR (255%) in all acquisitions (P < 0.025). IOSB and REX-DIR images were found to have different levels of SNR and CNR (P < 0.05) with IOSB values observed to be larger. Normalized to a common imaging time, the higher spatial resolution imaging at 3.0 T and the lower spatial resolution imaging at 1.5 T provided similar levels of wall-lumen CNR (P = NS). CONCLUSION: Multislice carotid wall imaging at 3.0 T with IOSB and REX-DIR benefits from improved SNR and CNR relative to 1.5 T, and allows for higher spatial resolution carotid artery wall imaging.     

Signal-to-noise ratio and spectral linewidth improvements between 1.5 and 7 Tesla in proton echo-planar spectroscopic imaging.

This study characterizes gains in sensitivity and spectral resolution of proton echo-planar spectroscopic imaging (PEPSI) with increasing magnetic field strength (B(0)). Signal-to-noise ratio (SNR) per unit volume and unit time, and intrinsic linewidth (LW) of N-acetyl-aspartate (NAA), creatine (Cr), and choline (Cho) were measured with PEPSI at 1.5, 3, 4, and 7 Tesla on scanners that shared a similar software and hardware platform, using circularly polarized (CP) and eight-channel phased-array (PA) head coils. Data were corrected for relaxation effects and processed with a time-domain matched filter (MF) adapted to each B(0). The SNR and LW measured with PEPSI were very similar to those measured with conventional point-resolved spectroscopy (PRESS) SI. Measurements with the CP coil demonstrated a nearly linear SNR gain with respect to B(0) in central brain regions. For the PA coil, the SNR-B(0) relationship was less than linear, but there was a substantial SNR increase in comparison to the CP coil. The LW in units of ppm decreased with B(0), resulting in improved spectral resolution. These studies using PEPSI demonstrated linear gains in SNR with respect to B(0), consistent with theoretical expectations, and a decrease in ppm LW with increasing B(0).     

High resolution spectroscopic imaging of GABA at 3 Tesla

A spin echo‐based MRSI sequence was developed to acquire edited spectra of γ‐aminobutyric acid in an entire slice. Water and lipid signals were suppressed by a dual‐band presaturation sequence, which included integrated outer volume suppression pulses for additional lipid suppression. Experiments in three normal volunteers were performed at 3 T using a 32‐channel head coil. High signal‐to‐noise ratio spectra and metabolic images of γ‐aminobutyric acid were acquired from nominal 4.5 cm³ voxels (estimated actual voxel size 7.0 cm³) in a scan time of 17 min. The sequence is also expected to co‐edit homocarnosine and macromolecules, giving a composite γ‐aminobutyric acid⁺ resonance. The γ‐aminobutyric acid⁺ to water ratio was measured using a companion water MRSI scan and was found to correlate linearly with the % gray matter (GM) of each voxel (γ‐aminobutyric acid⁺/water = (1.5 × GM + 3.2) × 10^?5, R = 0.27), with higher γ‐aminobutyric acid⁺ levels in gray matter compared with white. In conclusion, high signal‐to‐noise ratio γ‐aminobutyric acid‐MRSI is possible at 3 T within clinically feasible scan times. Magn Reson Med, 2011. © 2010 Wiley‐Liss, Inc.     

Myocardial triglycerides and systolic function in humans: in vivo evaluation by localized proton spectroscopy and cardiac imaging.

Recent experimental data suggest that adiposity directly damages the heart by promoting ectopic deposition of triglyceride, a process known as myocardial steatosis. The goal of this study was to develop and validate proton magnetic resonance spectroscopy ((1)H MRS) as an in vivo tool to measure myocardial lipid content. Complementary studies in rat tissue ex vivo and in 15 healthy humans in vivo provided evidence that (1)H MRS constitutes a reproducible technique for the measurement of myocardial triglyceride. In myocardial tissue from Zucker rats, the (1)H MRS measurement of triglyceride matched that obtained by biochemical measurement (P < 0.001). In human subjects triglyceride was evident in the hearts of even the very lean individuals and was elevated in overweight and obese subjects. Increased myocardial triglyceride content was accompanied by elevated LV mass and suppressed septal wall thickening as measured by cardiac imaging.     

Diffusion-weighted whole-body MR imaging with background body signal suppression: a feasibility study at 3.0 Tesla

The purpose was to provide a diffusion-weighted whole-body magnetic resonance (MR) imaging sequence with background body signal suppression (DWIBS) at 3.0 Tesla. A diffusion-weighted spin-echo echo-planar imaging sequence was combined with the following methods of fat suppression: short TI inversion recovery (STIR), spectral attenuated inversion recovery (SPAIR), and spectral presaturation by inversion recovery (SPIR). Optimized sequences were implemented on a 3.0- and a 1.5-Tesla system and evaluated in three healthy volunteers and six patients with various lesions in the neck, chest, and abdomen on the basis of reconstructed maximum intensity projection images. In one patient with metastases of malignant melanoma, DWIBS was compared with ¹⁸F-fluorodeoxyglucose positron emission tomography (FDG-PET). Good fat suppression for all regions and diagnostic image quality in all cases could be obtained at 3.0 Tesla with the STIR method. In comparison with 1.5 Tesla, DWIBS images at 3.0 Tesla were judged to provide a better lesion-to-bone tissue contrast. However, larger susceptibility-induced image distortions and signal intensity losses, stronger blurring artifacts, and more pronounced motion artifacts degraded the image quality at 3.0 Tesla. A good correlation was found between the metastases as depicted by DWIBS and those as visualized by FDG-PET. DWIBS is feasible at 3.0 Tesla with diagnostic image quality.     

High-resolution proton density weighted three-dimensional fast spin echo (3D-FSE) of the knee with IDEAL at 1.5 Tesla: comparison with 3D-FSE and 2D-FSE--initial experience

Purpose:

To assess the feasibility of combining three‐dimensional fast spin echo (3D‐FSE) and Iterative‐decomposition‐of water‐and‐fat‐with‐echo asymmetry‐and‐least‐squares‐estimation (IDEAL) at 1.5 Tesla (T), generating a high‐resolution 3D isotropic proton density‐weighted image set with and without “fat‐suppression” (FS) in a single acquisition, and to compare with 2D‐FSE and 3D‐FSE (without IDEAL).

Materials and Methods:

Ten asymptomatic volunteers prospectively underwent sagittal 3D‐FSE‐IDEAL, 3D‐FSE, and 2D‐FSE sequences at 1.5T (slice thickness [ST]: 0.8 mm, 0.8 mm, and 3.5 mm, respectively). 3D‐FSE and 2D‐FSE were repeated with frequency‐selective FS. Fluid, cartilage, and muscle signal‐to‐noise ratio (SNR) and fluid‐cartilage contrast‐to‐noise ratio (CNR) were compared among sequences. Three blinded reviewers independently scored quality of menisci/cartilage depiction for all sequences. “Fat‐suppression” was qualitatively scored and compared among sequences.

Results:

3D‐FSE‐IDEAL fluid‐cartilage CNR was higher than in 2D‐FSE (P < 0.05), not different from 3D‐FSE (P = 0.31). There was no significant difference in fluid SNR among sequences. 2D‐FSE cartilage SNR was higher than in 3D FSE‐IDEAL (P < 0.05), not different to 3D‐FSE (P = 0.059). 2D‐FSE muscle SNR was higher than in 3D‐FSE‐IDEAL (P < 0.05) and 3D‐FSE (P < 0.05). Good or excellent depiction of menisci/cartilage was achieved using 3D‐FSE‐IDEAL in the acquired sagittal and reformatted planes. Excellent, homogeneous “fat‐suppression” was achieved using 3D‐FSE‐IDEAL, superior to FS‐3D‐FSE and FS‐2D‐FSE (P < 0.05).

Conclusion:

3D FSE‐IDEAL is a feasible approach to acquire multiplanar images of diagnostic quality, both with and without homogeneous “fat‐suppression” from a single acquisition. J. Magn. Reson. Imaging 2012;361‐369. © 2011 Wiley Periodicals, Inc.     

Cross-sectional and in-plane coronary vessel wall imaging using a local inversion prepulse and spiral read-out: a comparison between 1.5 and 3 Tesla

Purpose:

To compare cross‐sectional and in‐plane coronary vessel wall imaging using a spiral readout at 1.5 and 3 Tesla (T).

Materials and Methods:

Free‐breathing coronary vessel wall imaging using a local inversion technique and spiral readout was implemented. Images were acquired in ten healthy adult subjects on a 3T clinical scanner using a 32‐element cardiac coil and repeated on a 1.5T clinical scanner using a 5‐element coil.

Results:

Cross‐sectional and in‐plane spiral vessel wall imaging was performed at both 1.5 and 3T. In cross‐sectional images, artifact scores were superior at 1.5T (P < 0.05) but no significant difference was found in image quality scores compared with 3T. Image quality (P < 0.01) and artifact scores (P < 0.01) were found to be superior for in‐plane images at 1.5T. Vessel wall sharpness in the in‐plane orientation was also found to be higher at 1.5T (P < 0.03).

Conclusion:

Although excellent in‐plane coronary vessel wall images can be acquired at 3T, the overall robustness may be affected by off‐resonance blurring due to increased B0 inhomogeneity compared with 1.5T. J. Magn. Reson. Imaging 2012;35:969–975. © 2011 Wiley Periodicals, Inc.     

Myocardial tagging with 3D-CSPAMM

PURPOSE: To introduce a true three-dimensional (3D) tagging technique for the assessment of myocardial tissue motion. MATERIALS AND METHODS: To generate a 3D tagging grid, a complementary spatial modulation of magnetization (CSPAMM) was applied in three spatial directions. Imaging was performed using a conventional fast 3D gradient-echo sequence. For automatic analysis of the 3D-CSPAMM data set, evaluation software, based on a 3D extension of the HARP technique, was used. RESULTS: Successful application of the 3D-CSPAMM technique in healthy subjects allowed the accurate determination of quantitative 3D motion patterns in the human heart. CONCLUSION: 3D-CSPAMM may contribute to the quantification of the local 3D myocardial motion pattern throughout the cardiac cycle.     

Quantitative analysis of regional left ventricular function after myocardial infarction in the pig assessed with cine magnetic resonance imaging

To assess the accuracy of quantitative analysis of global and regional wall motion and wall thickening of the left ventricle with cine magnetic resonance (MR), images obtained in eight pigs before and after myocardial infarction were compared with those obtained using gadolinium diethylenetriaminepen-taacetic acid (Gd-DTPA)-enhanced multislice spin-echo MR imaging and determination of pathology. The region with abnormal wall motion and wall thickening, as determined with cine MR imaging, identified the same region of infarction as indicated by Gd-DTPA-enhanced spin-echo MR imaging and pathology. Within the infarcted region wall motion and wall thickening analyzed with the centerline method were significantly reduced. We conclude that the use of quantitative analysis of cine MR images accurately determines localization and extent of regional left ventricular dysfunction in the infarcted heart in vivo. This analysis using dedicated software including the centerline method allows sequential assessment of regional left ventricular function in normal and infarcted hearts.     

MR liver imaging and cholangiography in the presence of surgical metallic clips at 1.5 and 3 Tesla

To evaluate whether clips from prior cholecystectomy impair image quality during magnetic resonance cholangiography (MRC) at 3 Tesla (T) compared with 1.5 T, surgical clips were embedded in a gel phantom and positioned at predefined distances from a fluid-filled tube designed to simulate the bile duct. The maximum clip distance was noted where susceptibility artifacts obscured the fluid-filled tube at 1.5 T and 3 T. Susceptibility artifact size was calculated for each sequence within each magnet class. In vivo analysis included 42 patients postcholecystectomy who underwent MRC at either 1.5 T or 3 T. In vitro, mean area of susceptibility artifacts was 104 mm² on 3-T and 75 mm² on 1.5-T MR imaging (MRI). While surgical clips within a 2-mm range impaired visualization of the fluid-filled tube on 1.5-T MRI, this range increased to 4 mm on 3-T MRI. In vivo, MRC image quality was impaired by susceptibility artifacts in three of 21 cases at 3 T and in two of 21 cases at 1.5 T. Overall, biliary pseudo-obstructions due to susceptibility artifacts from cholecystectomy surgical clips were not substantially more common on 3-T MRC in clinical practice, and patients with a history of prior cholecystectomy should not be excluded from a 3-T MRC.     

Parallel imaging of knee cartilage at 3 Tesla

PURPOSE: To evaluate the feasibility and reproducibility of quantitative cartilage imaging with parallel imaging at 3T and to determine the impact of the acceleration factor (AF) on morphological and relaxation measurements. MATERIALS AND METHODS: An eight-channel phased-array knee coil was employed for conventional and parallel imaging on a 3T scanner. The imaging protocol consisted of a T2-weighted fast spin echo (FSE), a 3D-spoiled gradient echo (SPGR), a custom 3D-SPGR T1rho, and a 3D-SPGR T2 sequence. Parallel imaging was performed with an array spatial sensitivity technique (ASSET). The left knees of six healthy volunteers were scanned with both conventional and parallel imaging (AF = 2). RESULTS: Morphological parameters and relaxation maps from parallel imaging methods (AF = 2) showed comparable results with conventional method. The intraclass correlation coefficient (ICC) of the two methods for cartilage volume, mean cartilage thickness, T1rho, and T2 were 0.999, 0.977, 0.964, and 0.969, respectively, while demonstrating excellent reproducibility. No significant measurement differences were found when AF reached 3 despite the low signal-to-noise ratio (SNR). CONCLUSION: The study demonstrated that parallel imaging can be applied to current knee cartilage quantification at AF = 2 without degrading measurement accuracy with good reproducibility while effectively reducing scan time. Shorter imaging times can be achieved with higher AF at the cost of SNR.     

磁共振电影成像估价心肌功能

采用磁共振电影成像（Ｃｉｎｅ－ＭＲＩ）估价９例正常人和１３例左心室陈旧心肌梗塞患者的左心室功能状况。其方法是应用小角度翻转角、短ＴＲ和信号聚焦的梯度场成像序列，通过观察左心室各段肌壁运动状态和测量肌壁厚度的变化以判断各段肌壁的运动功能。结果提示，Ｃｉｎｅ－ＭＲＩ无论对心肌壁运动状态的定性还是运动功能的定量均有一定的价值。     

3.0 Tesla vs 1.5 Tesla breast magnetic resonance imaging in newly diagnosed breast cancer patients

AIM:To compare 3.0 Tesla(T) vs 1.5T magnetic resonance(MR) imaging systems in newly diagnosed breast cancer patients.METHODS:Upon Institutional Review Board approval,a Health Insurance Portability and Accountability Actcompliant retrospective review of 147 consecutive 3.0T MR examinations and 98 consecutive 1.5T MR examinations in patients with newly diagnosed breast cancer between 7/2009 and 5/2010 was performed.Eleven patients who underwent neoadjuvant chemotherapy in the 3.0T group were excluded.Mammographically occult suspicious lesions(BIRADS Code 4 and 5) additional to the index cancer in the ipsilateral and contralateral breast were identified.Lesion characteristics and pathologic diagnoses were recorded,and results achieved with both systems compared.Statistical significance was analyzed using Fisher’s exact test.RESULTS:In the 3.0T group,206 suspicious lesions were identified in 55%(75/136) of patients and 96%(198/206) of these lesions were biopsied.In the 1.5T group,98 suspicious lesions were identified in 53%(52/98) of patients and 90%(88/98) of these lesions were biopsied.Biopsy results yielded additional malignancies in 24% of patients in the 3.0T group vs 14% of patients in the 1.5T group(33/136 vs 14/98,P = 0.07).Average size and histology of the additional cancers was comparable.Of patients who had a suspicious MR imaging study,additional cancers were found in 44% of patients in the 3.0T group vs 27% in the 1.5T group(33/75 vs 14/52,P = 0.06),yielding a higher positive predictive value(PPV) for biopsies performed with the 3.0T system.CONCLUSION:3.0T MR imaging detected more additional malignancies in patients with newly diagnosed breast cancer and yielded a higher PPV for biopsies performed with the 3.0T system.