Comparison of image quality and application values on different field-of-view diffusion-weighted imaging of breast cancer

Abstract

Background

Single-shot echo planar imaging (SS-EPI) diffusion-weighted magnetic resonance imaging (DW-MRI) is the most-widely sequence in breast MRI. MRI artifacts and magnetic susceptibility are sometimes severe when this sequence is utilized at 3T.

Purpose

To compare the imaging quality, ADC values between SS-EPI DWI sequence and two reduced field-of-view (rFOV) DWI sequences of breast cancer.

Material and Methods

Twelve cases with breast cancer were scanned using SS-EPI DWI (FOV, 360 × 360 mm), rFOV DWI1 (FOV, 360 × 180 mm), and rFOV DWI2 (FOV, 280 × 140 mm), respectively. Image quality (scores 1 to 5) and ADC values of breast imaging were compared among three groups by different DWI sequences. SNR were compared between rFOV DWI1 and rFOV DWI2.

Results

The imaging quality score of 12 cases was 5.00 in rFOV DWI1, 3.60 in SS-EPI DWI, and 3.75 in rFOV DWI2. The mean ADC value of 12 cases was 1.211 × 10⁻³mm²/s in SS-EPI DWI, 1.107 × 10^–3mm²/s in rFOV DWI1, and 1.038 × 10^–3mm²/s in rFOV DWI2. SNR of rFOV DWI1 images was much higher than that of rFOV DWI2.

Conclusion

rFOV DWI1 is the optimal DWI sequence in our study. Comparing with SS EPI DWI, suitable rFOV DWI has an obvious advantage, which can present higher image resolution and less distortion. It may be helpful in the diagnosis of breast cancer.

Keywords

Breast magnetic resonance imaging (MRI)diffusion-weighted imaging (DWI)apparent diffusion coefficient (ADC)

Introduction

Currently, single-shot echo planar imaging (SS-EPI) diffusion-weighted magnetic resonance imaging (DW-MRI) is the most widely used sequence for breast MRI due to its high speed and efficiency. However, its image resolution is greatly limited by MRI artifacts and magnetic susceptibility. In some cases, artifacts and image distortions will also lead to unavailable images for diagnosis or the detection of small foci (1 –6). Reduced field-of-view diffusion-weighted imaging (rFOV DWI) employs two-dimensional radiofrequency (2DRF) just to excite small ranges of region of interest (ROI), which not only help diminish artifacts but also obtain DW images of better quality and higher resolution. This technique has been most frequently used in mapping the spine and brain (7 –11), but less often in breast studies (12,13). To our knowledge, no reports have been presented concerning FOV optimization in breast imaging.

In this study, 12 clinical patients underwent a 3.0 T MRI examination using SS-EPI DWI, as well as two rFOV DWI sequences. We then compared the image quality, ADC values, and clinical values of SS-EPI DWI and two rFOV DWI sequences to evaluate the optimal DWI sequence.

Material and Methods

Patient data

MRI was performed in 12 female patients (age range, 26–65 years; mean age, 46.6 years) with suspicious breast masses enrolled at our hospital from August 2012 to February 2014. All 12 cases were indicated to be breast cancer by MRI, and confirmed malignant by histopathological studies. Among them, seven cases were infiltrative ductal carcinoma, four cases ductal carcinoma in situ, and one case was an invasive micropapillary carcinoma. Longest diameter measured on MR images were in the range of 0.8–4.6 cm (mean, 2.5 cm). Patients that had undergone previous breast surgery, breast biopsy, and breast cancer chemotherapy were excluded. Patients who had breast masses without histopathologic confirmation were also excluded. This study was authorized by the hospital ethics committee. All patients were informed and gave a signed written consent.

Data acquisition

MRI scanning was performed by a 3.0-T MR (Discovery750, GE Healthcare, Waukesha, WI, USA), with an 8-channel dedicated breast coil. Patients lay in prone position, with the breasts suspended within the breast coil.

MR unenhanced scanning

After the conventional three-plane positioning scanning, cross-section fast spin echo (FSE) sequence T1-weighted (T1W) imaging (repetition time [TR], 1500 ms; echo time [TE], 25 ms; FOV, 360 × 360 mm; matrix, 320 × 256; layer thickness, 4 mm; gap, 1 mm; number of signals acquired [NEX], 2) and Iterative Dixon water-fat separation with Echo Asymmetry and Least-squares estimation (IDEAL) fat suppression T2-weighted (T2W) imaging (TR, 2580 ms; TE, 85 ms; FOV, 360 × 360 mm; matrix, 320 × 256, layer thickness, 4 mm; gap, 1 mm; bandwith [BW], 62.5 kHz) were performed.

The scanning parameters of SS-EPI DWI were: TR, 3600 ms; TE, 78 ms; FOV, 360 × 360 mm; matrix, 160 × 160; slice thickness, 4 mm; gap, 1 mm; BW, 250 kHz; NEX, 6; b value, 800 s/mm². The scanning time was 76 s. Before the scanning, a combined sequence with array spatial sensitivity encoding technique (ASSET) was conducted.

The scanning parameters of rFOV DWI1 were: TR, 3600 ms; TE, 62 ms; FOV, 360 × 180 mm; matrix, 160 × 160; slice thickness, 4 mm; gap, 1 mm; BW, 250 kHz; NEX, 6; b value, 800 s/mm². The scanning time was 164 s.

The scanning parameters of rFOV DWI2 were: TR, 3600 ms; TE, 59 ms; FOV, 280 × 140 mm; matrix, 160 × 160; slice thickness, 4 mm; gap, 1 mm; BW, 250 kHz; NEX, 6; b value, 800 s/mm². The scanning time was 164 s.

Dynamic contrast-enhanced MRI

For dynamic contrast-enhanced (DCE) MRI, Gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA) (Magnevist, Bayer Schering Pharma AG, Berlin, Germany) was injected as a bolus at the dose of 0.1 mmol/kg body weight and a flow rate of 2 mL/s, followed by 20 mL of saline at the same rate to rinse the residual contrast agent in the catheter. Before the injection, a mask was scanned. After the injection, six time phases of volume imaging for breast assessment (VIBRANT-Flex) were continuously acquired with about 50 s per phases, 60 s interval between the fifth and sixth phase, and a total scanning of 410 s (TR, 3.9 ms; TE, 1.7 ms; flip angle, 12°; FOV, 360 × 360 mm; matrix, 320 × 320; and layer thickness, 1.4 mm).

Image analysis

All image analyses were carried out on ADW4.5 workstation (GE Healthcare) by two senior radiologists with over 10 years of experience in breast MRI. The radiologists were blinded to sequence type and evaluated the image quality together. SS-EPI DWI, rFOV DWI1, and rFOV DWI2 images were referred to as sequence 1, sequence 2, and sequence 3, respectively. Sequence number varied from case to case to improve blinding, but complete blinding for sequence type was not possible because of the visible FOV differences among the three sequences images.

Assessment of imaging quality

According to the susceptibility artifacts, chemical shift artifacts, and image distortion on the display of anatomical details, the images were scored as level 1 to 5 where 1 was non-diagnostic, images cannot be used in diagnosis; 2 was poor, lots of artifacts influence diagnosis but not all anatomical structures; 3 was moderate, a few magnetic susceptible artifacts or opacities, and reconstructed artifacts, but the diagnoses based on MR breast images are trustable; 4 was good, with comparatively clear anatomical details, as well as few susceptibility artifacts and reconstructed artifacts; 5 was excellent, with distinct anatomical details and no artifacts .

Measurement of ADC values, signal intensity values, and noise values

Circular regions of interest (ROIs) in the range o of 10–20 mm² were used to measure breast ADC values and signal intensity values of the three sequences as well as noise values in both rFOV DWI sequence. When ADC values and signal intensity values of normal glands was measured, a nipple slice with plenty of glandular tissues was selected in order to avoid artifacts. On each slice, three ROIs of bilateral breasts were randomly chosen to calculate the mean value (the three DWI sequence positions were consistent). When ADC value and signal intensity value of a breast cancer was detected, the substantial part with the greatest slice of pathologic changes was selected, avoiding cystic necrosis and bleeding areas, and then the mean value of three ROIs was taken. When noise value was measured, three ROIs were randomly placed in the blank area of two rFOV DWI images to calculate the mean value (the two rFOV DWI sequence positions were consistent).

Statistical analysis

The comparison of image quality of three sequences was conducted by non-parametric Friedman test; if the difference was statistically significant, then we compared two of three sequences using the Wilcoxon paired signed rank test. The comparison of SNR values of two rFOV DWI was conducted by paired t test. ADC values differences of three sequences were compared by single factor variance analysis. P < 0.05 refers to a significant difference, while P < 0.01 means the difference is extremely significant.

Results

The comparison of image quality of three sequences

The images obtained by SS-EPI DWI present deformation, distortion, chemical shift artifacts, and susceptibility artifacts to variable degrees, while there was no deformation, distortion, or obvious artifacts by rFOV DWI (Figs. 1 and 2). Based on the formula image resolution = FOV/matrix, the image resolution of SS-EPI DWI, rFOV DWI1, and rFOV DWI2 was 2.25 × 2.25 mm ([360 × 360 mm] / [160 × 160]), 2.25 × 1.225 mm ([360 × 180 mm] / [160 × 160]), and 1.75 × 0.875 mm ([280 × 140 mm] / [160 × 160]), respectively. The image resolution of rFOV DWI1 was the best in three sequences. The imaging quality score of the three sequences is given in Table 1.

Fig. 1.

Imaging results from a 34-year-old woman with infiltrative ductal carcinoma in the left breast. (a) SS-EPI DWI shows a few chemical shift artifacts and susceptibility artifacts, with slight distortion and clear details of left focus, scored 3. (b) rFOV DWI1 show the focus with clear anatomic details, no significant artifact nor distortion, so the image quality scores is 5. (c) rFOV DWI2 shows the focus with clear details and no significant artifact nor deformation, nipples are not included, scored 4. (d) DCE image.

Fig. 2.

Imaging results from a 26-year-old woman with ductal carcinoma in situ of the right breast. (a) SS-EPI DWI shows the focus with a moderately strong signal, blurred details, and a few artifacts and distortion, so the image quality score is 3. (b) rFOV DWI1 of the focus shows a strong signal with clear anatomic details, no artifacts, and the image quality score is 5. (c) rFOV DWI2 of the focus shows a medium high signal with comparatively clear anatomic details and increased noise, no artifacts, the image quality score is 4. (d) DCE image.

Table 1.

The scores of three sequences of 12 cases.

Score	SS-EPI DWI	rFOV DWI1	rFOV DWI2
1	0	0	0
2	0	0	0
3	3	0	1
4	9	0	11
5	0	12	0
Average score	3.60	5	3.75
X² value P value	21.800 0.0001

Comparison between SS-EPI DWI and rFOV DWI1: Z = −3.357, P = 0.001.

Comparison between rFOV DWI1 and rFOV DWI2: Z = −3.217, P = 0.001.

Comparison between SS-EPI DWI and rFOV DWI2: Z = −1.000, P = 0.357.

Comparison of SNR of rFOV DWI1 and rFOV DWI2

The SNR of rFOV DWI1 was 112.291 ± 46.553, and that of rFOV DWI2 was 65.604 ± 29.293, the difference was of great statistically significance (t = 5.978, P = 0.001).

Comparison of ADC values of the three sequences

The mean ADC value of the 12 cases was 1.211 × 10⁻³mm²/s (SD, 0.226) in SS-EPI DWI, 1.107 × 10^–3mm²/s (SD, 0.181) in rFOV DWI1, and 1.038 × 10⁻³mm²/s (SD, 0.191) in rFOV DWI2. The ADC values of breast cancer has different degrees of decline as FOV is reduced, but the difference was not statistically significant (F = 2.236, P = 0.123).

Discussion

Currently, single-shot EPI (SS-EPI) is the most frequently used DWI method in clinical practice. However, EPI sequence is limited by a long TE as well as narrow bandwidth in the direction of phase encoding and is likely to produce susceptibility and chemical shift artifacts, leading to low image quality, as well as low SNR and resolution. Array spatial sensitivity encoding technique (ASSET) can not only accelerate scanning and reduce acquisition times, and accordingly shorten examination time, but also diminish motion artifacts, blurred images and distortion, which will improve imaging quality of breast DWI. However, the advantage of ASSET is also limited. Due to the interlacing acquisition of K space, the imaging SNR would decrease as the speed increases. As a consequence, the MR unit will automatically remove the noise of the blank areas during ASSET imaging reconstruction so as to increase SNR (1,6,13 –18). Because the noise of ASSET cannot be detected in the blank area, this study only compared SNR of rFOV DWI1 and rFOV DWI2.

rFOV DWI uses two-dimensional selective excitation frequency technology to achieve strip excitation during slice selection excitation. It has no radiofrequency cross talk among slices in multi-scanning, thus decreases the readout of phase encoding (PE) line and TE time. Comparatively it increases the BW in the direction of PE. It can also decrease image deformation and artifacts. More importantly, it decreases FOV while maintaining a large acquisition matrix, so the resolution is increased (1,10 –12,19,20). ASSET shortens scanning time by increasing the acquisition gap of k-space (i.e. decreasing the sampling steps of PE), whereas rFOV DWI obtains images of high spatial resolution through decreased sampling steps of PE (21,22). If rFOV DWI adopts ASSET simultaneously, SNR will have a marked drop, indicating that current rFOV DWI sequences are not compatible with ASSET, so the scanning time of the rFOV DWI sequence is much longer than the conventional SS-EPI DWI sequence.

In this study, the FOV of rFOV DWI1 was 360 mm in the left-to-right direction (frequency-encoding direction) and 180 mm in the fore-and-aft direction (PE direction), which includes the bilateral breast glands completely. Furthermore, in the reduced vision field in the fore-and-aft direction, there are fewer motion artifacts in breast tissue caused by heartbeat and lung motion (11). The FOV of rFOV DWI2 is 280 mm in the left-to-right direction (frequency-encoding direction) and 140 mm in the fore-and-aft direction (PE direction), which cannot include larger breasts.

Zaharchuk et al. (7) applied rFOV DWI to the studies of spine and spinal cord lesions and found that rFOV DWI has high spatial resolution, high SNR, and less deformation, which conduces to spine and spinal cord lesion diagnosis. Singer et al. (11) applied rFOV DWI and SS-EPI DWI on progressive breast cancer. They found that rFOV DWI cannot only increase image resolution, but also decrease susceptibility artifacts and chemical shift artifacts. So they concluded that this sequence is of great value in assessing response of neoadjuvant chemotherapy of breast cancer.

SS-EPI DWI and rFOV DWI in this study uses the same parameters of b value, matrix, slice thickness, BW, and TR. Image quality scores indicate that, even when ASSET has been used in SS-EPI DWI to shorten scanning time and improve image quality, the rFOV DWI1 still has a remarkably better image quality and less deformation than the SS-EPI DWI and rFOV DWI2. Hence, suitable rFOV DWI is better at assessing restricted breast lesion diffusion.

However, compared with SS-EPI DWI, rFOV DWI also has shortcomings. Smaller FOV causes obviously lower SNR. rFOV DWI2 has the highest spatial resolution in this study, but the image quality was poorer than rFOV DWI1, indicating that smaller FOV is not necessarily associated with better image quality. Moreover, small FOV could not include larger breasts.

The differences between average ADC values of the three sequences in breast cancer are not statistically significant, which probably can be caused by small samples. Singer et al. (11) considered that rFOV DWI reduces the partial volume effect between tumors and normal breast gland tissue, so the ADC value of breast cancer detected by rFOV DWI is lower than of SS-EPI DWI being possibly a more accurate ADC value (23 –26).

Our study has several limitations. First, only 12 cases of breast cancer were included. The number of cancer cases was relatively small in comparison with prior studies, and as the histological diagnosis varied, the ADC value may be affected. More lesions overall, such as cysts, incidental benign masses, will be included in future studies to achieve better results. Second, the scanning time of rFOV DWI is longer than that of SS-EPI. Finally, the study is just a single-center study.

In conclusion, rFOV DWI has higher resolution and less distortion than SS-EPI DWI. Suitable rFOV DWI may be helpful in the diagnosis of breast cancers. However, too small FOV will cause lower SNR and influence the display of lesions.

Footnotes

Conflict of interest

None declared.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Partridge

McDonald

. Diffusion weighted magnetic resonance imaging of the breast: protocol optimization, interpretation, and clinical applications. Magn Reson Imaging Clin N Am 2013; 21: 601–624.

Wenkel

Geppert

Schulz-Wendtland

. Diffusion weighted imaging in breast MRI: comparison of two different pulse sequences. Acad Radiol 2007; 14: 1077–1083.

Baltzer

Renz

Herrmann

. Diffusion-weighted imaging (DWI) in MR mammography (MRM): clinical comparison of echo planar imaging (EPI) and half-Fourier single-shot turbo spin echo (HASTE) diffusion techniques. Eur Radiol 2009; 19: 1612–1620.

Le Bihan

Poupon

Amadon

. Artifacts and pitfalls in diffusion MRI. J Magn Reson Imaging 2006; 24: 478–488.

Bammer

Keeling

Augustin

. Improved diffusion-weighted single-shot echo-planar imaging (EPI) in stroke using sensitivity encoding (SENSE). Magn Reson Med 2001; 46: 548–554.

Kuhl

Gieseke

von Falkenhausen

. Sensitivity encoding for diffusion-weighted MR imaging at 3.0 T: intraindividual comparative study. Radiology 2005; 234: 517–526.

Zaharchuk

Saritas

Andre

. Reduced field-of-view diffusion imaging of the human spinal cord: comparison with conventional single-shot echo-planar imaging. Am J Neuroradiol 2011; 32: 813–820.

Wilm

Svensson

Henning

. Reduced field-of-view MRI using outer volume suppression for spinal cord diffusion imaging. Magn Reson Med 2007; 57: 625–630.

von Morze

Kelley

Shepherda

. Reduced field-of-view diffusion-weighted imaging of the brain at 7 T. Magn Reson Imaging 2010; 28: 1541–1545.

10.

Karampinos

Van

Olivero

. High resolution reduced-FOV diffusion tensor imaging of the human pons with multi-shot variable density spiral at 3T. Conf Proc IEEE Eng Med Biol Soc 2008, pp. 5761–5764.

11.

Singer

Wilmes

Saritas

. High-resolution diffusion-weighted magnetic resonance imaging in patients with locally advanced breast cancer. Acad Radiol 2012; 19: 526–534.

12.

Wilmes

McLaughlin

Newitt

. High-resolution diffusion-weighted imaging for monitoring breast cancer treatment response. Acad Radiol 2013; 20: 581–589.

13.

Blaimer

Breuer

Mueller

. SMASH, SENSE, PILS, GRAPPA: how to choose the optimal method. Top Magn Reson Imaging 2004; 15: 223–236.

14.

Kuroki

Nasu

Kuroki

. Diffusion-weighted imaging of breast cancer with the sensitivity encoding technique: analysis of the apparent diffusion coefficient value. Magn Reson Med Sci 2004; 3: 79–85.

15.

Friedman

Swaminathan

Smith

. SENSE imaging of the breast. Am J Roentgenol 2005; 184: 448–451.

16.

Brink

Watanabe

Kuhl

. Implication of SENSE MR in routine clinical practice. Eur J Radiol 2003; 46: 3–27.

17.

Weiger

Pruessmann

Boesiger

. Cardiac real-time imaging using SENSE. Magn Reson Med 2000; 43: 177–184.

18.

Weiger

Pruessmann

Leussler

. Coil design for SENSE: a six-element cardiac array. Magn Reson Med 2001; 45: 495–504.

19.

Saritas

Cunningham

Lee

. DWI of the spinal cord with reduced FOV single-shot EPI. Magn Reson Med 2008; 60: 468–473.

20.

Jeong

Kim

Guo

. High-resolution DTI with 2D interleaved multislice reduced FOV single-shot diffusion-weighted EPI (2D ss-rFOV-DWEPI). Magn Reson Med 2005; 54: 1575–1579.

21.

Woodhams

Ramadan

Stanwell

. Diffusion-weighted imaging of the breast: principles and clinical applications. Radiographics 2011; 31: 1059–1084.

22.

Pereira

Martins

Figueiredo

. Assessment of breast lesions with diffusion-weighted MRI: comparing the use of different b values. Am J Roentgenol 2009; 193: 1030–1035.

23.

Peters

Vincken

van den Bosch

. Quantitative diffusion weighted imaging for differentiation of benign and malignant breast lesions: the influence of the choice of b-values. J Magn Reson Imaging 2010; 31: 1100–1105.

24.

Bogner

Gruber

Pinker

. Diffusion weighted MR for differentiation of breast lesions at 3.0 T: how does selection of diffusion protocols affect diagnosis? Radiology 2009; 253: 341–351.

25.

Tsushima

Takahashi-Taketomi

Endo

. Magnetic resonance (MR) differential diagnosis of breast tumors using apparent diffusion coefficient (ADC) on 1.5-T. J Magn Reson Imaging 2009; 30: 249–255.

26.

Kuroki

Nasu

. Advances in breast MRI: diffusion-weighted imaging of the breast. Breast Cancer 2008; 15: 212–217.