Megumi TAKECHI¹, Takaharu TSUDA¹, Hiroaki TANAKA¹, Shinji YOSHIOKA², Michinobu NAGAO³, and Teruhito MOCHIZUKI^1
1Department of Radiology, Ehime University School of Medicine, Shitsukawa,Toon, Ehime, Japan, ²Department of Radiology, Matsuyama Redcross Hospital, Japan,³Department of Molecular Imaging and Diagnosis, Kyushu University School of Medicine, Japan

Signal intensity of hepatic nodules detected by gadoxetic acid-enhanced MR imaging: Correlation with arterial and portal blood supply.

Eun-Suk Cho¹, and Jeong-Sik Yu^1
1Radiology, Yonsei University College of Medicine, Gangnam Severance Hospital, Seoul, Korea, Republic of

Gadoxetic acid-enhanced MRI improves detection of hepatocellular carcinoma (HCC). However, since the degree of liver enhancement in hepatobiliary phase is known to be lower in patients with chronic liver dysfunction, diagnostic value of hepatobiliary phase imaging of Gadoxetic acid-enhanced MRI might be expected to be lower in the diagnosis of HCC in patients with impaired liver function. Therefore, we tried to evaluate the value of hepatobiliary phase of gadoxetic acid-enhanced MRI in the diagnosis of HCC in patients with impaired liver function.

Tomoko Hyodo^1,2, Masahiro Okada¹, Yuki Kagawa¹, Sachiyo Kogita³, Seishi Kumano¹, Izumi Imaoka¹, Masatoshi Hori⁴, Kazunari Ishii¹, Yasuharu Imai³, Teruhito Mochizuki², Masatoshi Kudo⁵, and Takamichi Murakami^1
1Radiology, Kinki University Faculty of Medicine, Osaka-Sayama, Osaka, Japan, ²Diagnostic and Therapeutic Radiology, Ehime University Graduate School of Medicine, Toon, Ehime, Japan, ³Gastroenterology, Ikeda Municipal Hospital, Ikeda, Osaka, Japan, ⁴Radiology, Osaka University Graduate School of Medicine, Suita, Osaka, Japan,⁵Gastroenterology and Hepatology, Kinki University Faculty of Medicine, Osaka-Sayama, Osaka, Japan

To clarify the significance of hypovascular nodules presented with hypointensity on hepatobiliary phase of Gd-EOB-DTPA enhanced MRI in screening for hepatocellular carcinoma (HCC) in the cirrhotic liver, 122 hypovascular hepatic nodules with hypointensity on hepatobiliary phase in 54 patients were reviewed. The incidence rate of developing hypervascular HCC from such nodules was 32%. Patients with higher growth rate (GR) may have potential arterial hypervascularization, and GR of >2.8x10^-3(tumor volume doubling time of < 355 days) may justifiy follow-up at short intervals.

Hana Kim¹, Jeong-Sik Yu¹, Eun-Suk Cho¹, Jae-Joon Chung¹, Joo Hee Kim¹, and Ki Whang Kim^1
1Radiology, Yonsei University College of Medicine, Gangnam Severance Hospital, Seoul, Korea, Republic of

SPIO-enhanced DWI showed higher sensitivity in the detection of liver metastases

Mustafa R Bashir¹, Rajan T Gupta¹, Matthew S Davenport¹, Brian C Allen¹, Lisa M Ho¹, Daniel T Boll¹, and Elmar M Merkle^1
1Radiology, Duke University Medical Center, Durham, NC, United States

The purpose of this study is to evaluate whether the hepatocyte phase (delayed phase) image data set improves the detection and reader confidence in characterization of hepatocellular carcinoma in the MRI of the cirrhotic liver with Gd-EOB-DTPA.

Chen-Te Chou¹, and Ran-Chou Chen^2
1Radiology, Changhua Christian Hospital, Chang-Hua, Taiwan, Taiwan, ²Radiology, Taipei City Hospital, Taipei, Taiwan

Purpose: To evaluate the utility of gadoxetic acid-enhanced hepatocyte-phase imaging (HP) in characterization of T1W hyperintense nodules. Materials and Methods: 34 T1W hyperintense nodules in 19 patients with histopathological confirmation were included. ROC analysis was used to evaluate the diagnostic performance. Results: The mean size of dysplastic nodules were smaller than that of HCCs. (p<0.001) There were seven additional HCCs diagnosed using hepatocyte-phase imaging compared to conventional HCC diagnostic criteria (p=0.02). Conclusion: gadoxetic acid-enhanced MRI with HP is superior to conventional criteria alone in characterization of T1W hyperintense nodule.

David H Gultekin¹, Lawrence H Schwartz², Nancy E Kemeny³, Mithat Gonen⁴, Michael I D'Angelica⁵, Peter J Allen⁵, Yuman Fong⁵, Leslie H Blumgart⁵, Ronald P Dematteo⁵, and William R Jarnagin^5
1Radiology, Memorial Sloan-Kettering Cancer Center, New York, NY, United States, ²Radiology, Columbia University Medical Center, New York, NY, United States,³Medicine, Memorial Sloan-Kettering Cancer Center, New York, NY, United States, ⁴Epidemiology-Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, New York, United States, ⁵Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY, United States

The role of DCE-MRI and DWI-MRI have been evaluated in a Phase II study for the assessment of response to therapy in patients with unresectable primary liver cancers, hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC), undergoing regional combination chemotherapy and anti-angiogenic therapy through continuous hepatic arterial infusion (HAI) treatment procedure.

Yin Huang¹, Ethan Brodsky¹, Kevin Johnson¹, Eric Bultman², Debra Horng^1,3, Sean Fain^1,3, and Scott Reeder^1,3
1Medical Physics, University of Wisconsin Madison, Madison, WI, United States, ²Biomedical Engineering, University of Wisconsin Madison, Madison, WI, United States,³Radiology, University of Wisconsin Madison, Madison, WI, United States

A novel method for acquiring dynamic contrast-enhanced (DCE) MRI of liver perfusion with full abdominal coverage is used to measure the blood supply to liver parenchyma and focal liver lesions non-invasively. This study aims to provide semi-quantitative metrics to evaluate 3D regional perfusion in a feasibility study to measure perfusion in focal liver lesions. The method is designed to predict long-term tumor response to chemotherapy by providing quantitative measures of early changes in perfusion specific to liver tumors.

Xiuzhong Yao¹, Mengsu Zeng¹, He Wang², Fei Sun², Shengxiang Rao¹, and Yuan Ji^3
1Radiology, Zhongshan Hospital of Fudan University, Shanghai, Shanghai, China, People's Republic of, ²The applied science lab,GE Healthcare, ³Pathology, Zhongshan Hospital of Fudan University, Shanghai, Shanghai, China, People's Republic of

To evaluate 3.0T DWI technique on Normal pancreas and pancreatic cancer, five DWI acquisitions were all performed in 15 normal volunteers and 30 patients with pancreatic cancer. Artifacts, SNR and ADC value of normal pancreas and C, CNR and ADC value of pancreatic cancer were investigated. Statistically higher C and CNR of pancreatic cancer was noticed in respiratory-triggered DWI with MPG pulses in X¡¢Y¡¢Z direction and inversion recovery for fat saturation compared to other four DWI acquisitions, and its ADC value can better disclose histopathological state in pancreatic cancer, adjacent pancreatic tissue and distal pancreatitis.

Masoom A Haider¹, Wigdan Al-Sukhni², Kartik S Jhaveri¹, Heidi Rothenmund², Spring Holter², Steven Narod³, Malcolm Moore⁴, Stephanie Wilson⁵, and Steven Gallinger^2
1Medical Imaging, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada, ²Department of Surgery, University Health Network, University of Toronto, Toronto, Ontario, Canada, ³Women's College Research Institute, University of Toronto, Toronto, Ontario, Canada, ⁴Department of Medicine, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada, ⁵Medical Imaging, University Health Network, University of Toronto, Toronto, Ontario, Canada

High risk patients were screened annually with non-contrast 20 minute MRI protocol from 2003-2010. Pancreatic cancer was found in 2 of the 259 patients (0.7%): 57 yo female with a 1.5cm adenocarcinoma and a concomitant IPMN, stage IIB ; 80 yo male with unresectable pancreatic adenocarcinoma. IPMNs were seen in 14 patients (5.4%).Seven significant incidental findings (2.7%) were detected by MRI. MRI has the potential to detect pancreatic cancer in a high risk screening populations but frequent scanning and IV contrast and may be required to impact survival. IPMN’s are a common finding.

Santosh Bhandari¹, N. Jane Taylor², Ashish Sinha¹, J. James Stirling², Ian C Simcock², Arun Gupta¹, Robin K.S. Phillips¹, Susan K Clark¹, and Vicky J Goh^2
1Polyposis Registry, St Mark’s Hospital, London, United Kingdom, ²Paul Strickland Scanner Centre, Mount Vernon Hospital, Northwood, Middlesex HA6 2RN, United Kingdom

Familial adenomatous polyposis (FAP) is an autosomal dominant hereditary condition characterised by multiple colonic polyps at an early age, which if left untreated become cancerous. Extracolonic manifestations are now the leading cause of death following prophylactic colectomy. Up to a quarter of patients will develop desmoid tumours, arising from the musculo-aponeurotic tissues of which about 10% will grow relentlessly, resulting in death. Outcome cannot be reliably predicted from histologic findings. DCE-MRI was performed on desmoid tumours to assess their vascularisation and to explore if it is possible to use it to predict aggressive phenotypes.

Stefano Viotti¹, Giuseppe Petralia¹, Paul Eugene Summers¹, Luke Bonello¹, Moreno Pasin¹, Roberto Di Filippi¹, and Massimo Bellomi^1,2
1European Institute of Oncology, Milano, Italy, ²School of Radiology, Università Statale degli Studi di Milano, Milano, Italy

The aim of our study was to compare DCE-MR and pCT parameters, pre and post neoadjuvant chemoradiation therapy (NACRT) in patients with advanced rectal cancer. 13 patients underwent DCE-MR and pCT, before and after NACRT. Pre-treatment, moderate correlation was seen between Kep and BV (R=0.64), and trends between Ktrans and BV (R=0.56) and between Ve and PS (R= 0.55). Post-treatment Ktrans, Kep and IAUC60 had moderate correlations with PS (R=0.63, 0.61, 0.61 respectively). The changing relationships following NACRT between DCE and pCT parameters may relate to different models applied or balances between flow and permeability limited conditions for different contrast agents used.

Thiele Kobus¹, Andreas K Bitz², Mark J Van Uden¹, Miriam W Lagemaat¹, Stephan Orzada², Arend Heerschap¹, and Tom W.J. Scheenen^1,2
1Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, Gelderland, Netherlands, ²Erwin L. Hahn Institute for Magnetic Resonance Imaging, Essen, Germany

A thorough safety validation was performed to enable combined use of a ³¹P TxRx endorectal coil with a 8-channel ¹H TxRx array coil. The maximum RF power with the ³¹P coil was determined. The coupling between the coils while transmitting with the 8-channel ¹H coil was studied by B1⁺ and temperature mapping. Although an increased proton flip angle was observed near the endorectal coil wire, no local temperature increase was observed. The in vivo measurement using adiabatic excitation showed good quality of both prostate imaging and ³¹P spectra, demonstrating the potential of ³¹P MRSI of the prostate at high field.

Kirsten Margrete Selnæs¹, Ingrid Susanne Gribbestad¹, Helena Bertilsson^2,3, Alan Wright⁴, Anders Angelsen³, Arend Heerschap⁴, and May-Britt Tessem^1
1Department of Circulation and Medical Imaging, NTNU, Trondheim, Norway, ²Department of Laboratory Medicine and Children's and Women's Health, NTNU, Trondheim, Norway, ³Department of Urology, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway, ⁴Department of Radiology, Radboud University Nijmegen Medical Centre, Netherlands

Both in vivo and ex vivo spectroscopy are important in prostate cancer handling, with elevated (Choline+Creatine)/Citrate indicative of cancer. So far no studies have investigated the correlation of in vivo and ex vivo measurements in matched regions of the same prostate cancer patient. In this study ex vivo HR MAS spectra were compared to in vivo MRS spectra in spatially matched regions in the same prostate cancer patient. CC/C ratios from in vivo MRS and ex vivo HR MAS were significantly correlated with r=0.63. Both in vivo and ex vivo CC/C were significantly correlated to Gleason grade (r=0.62 and 0.65)

Alan James Wright¹, and Arend Heerschap^1
1Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands

Currently, prostate 3D MRSI acquisition methods have matured considerably, but automatic spectral processing remains a bottleneck to realise routine clinical application as processing is often hampered by variations in phase and chemical shift of metabolite peaks across the MRSI data set. Furthermore, the citrate resonance at 2.6ppm is sensitive to the salt concentration and pH of the local environment and can vary quite dramatically within one 3D MRSI slice. We have developed a phase and frequency alignment algorithm, specific to the unique problem of prostate cancer MRSI data sets, and tested it with simulated and patient data.

Alan James Wright¹, and Arend Heerschap^1
1Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands

MRSI is a promising technique for the detection and localization of tumours in patients with prostate cancer, the clinical relevance of this data may be realised by the application of Pattern recognition techniques. The performance of such algorithms is partly dependent on artifacts such as residual large lipid signals from the tissue surrounding the prostate; these signals are broad and can overlap with the citrate resonances at 2.6ppm that play a key role in cancer diagnosis by MRSI. We present an algorithm that uses prior knowledge to remove lipid signals from prostate MRSI data sets.

Yuriko Suzuki¹, Yoshifumi Kuroki², and Marc Van Cauteren^1
1MR Clinical Science, Philips Electronics Japan, Minato-ku, Tokyo, Japan, ²Tochigi Cancer Center, Utsunomiya, Tochigi, Japan

It is important to assess the response to neoadjuvant chemotherapy for optimizing the therapy regime. Recently, in vivo proton MR spectroscopy of breast is demonstrating promising results in early assessment of response to NAC. However those methods still have limitation in quantification. A new method is presented to measure tCho concentration using internal reference more accurately, by applying the correction of tissue water content using external reference. The changes in tCho following neoadjuvant chemotherapy significantly differed between responder group and non-responder group. Our data showed that tCho is a better predictor of early response to neoadjuvant chemotherapy.

Xiufeng Li¹, Chaitanya Kalavagunta¹, Michael T Nelson², and Greg J Metzger^1
1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States, ²Diagnostic Radiology, University of Minnesota, Minneapolis, MN, United States

The use of an endorectal coil can improve prostate perfusion imaging with arterial spin labeling methods by providing high signal noise ratios, decreased prostate motion, and reduced B0 distortions and/or signal loss. To characterize the noise and evaluate achievable resolutions, prostate perfusion studies were performed at 3T with an endorectal coil. These preliminary results show that high quality perfusion imaging maps can be achieved at 3T with an ERC. Physiological noise was found to be the major source of temporal error, and increased with imaging resolution. With more averages, temporal error can be limited to a relatively low level.

Xiufeng Li¹, Chaitanya Kalavagunta¹, and Greg Metzger^1
1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States

Non-contrast enhanced perfusion imaging method using arterial spin labeling can benefit prostate cancer patients with contraindications to the use of MR contrast-agent, and should be more suitable for repeated studies monitoring tumor progression or treatment response. To evaluate the feasibility of prostate perfusion using FAIR at 3T, multiple inversion perfusion studies were performed with healthy controls using surface array coils. Signal changes originating from labeled blood are measurable and correlate well with other measurement of prostate perfusion. However, SNR and physiological motion are limiting factors for obtaining perfusion maps, making the proposed method and hardware impractical for clinical applications. To our knowledge, these are the first attempts at obtaining ASL results in the human prostate.

Fiona M Fennessy¹, Sandeep N Gupta², Andriy Fedorov¹, Robert Mulkern¹, Yi Tang¹, Felipe Franco¹, Kemal Tuncali¹, Ehud Schmidt¹, and Clare Tempany^1
1Brigham and Women's Hospital, Boston, MA, United States, ²Functional Imaging Lab, GE Global Research Center, Niskayuna, NY, United States

Pharmacokinetic (PK) analysis allows for quantification of DCE MRI data in prostate cancer. The objective of this work is to determine the variability in PK analysis using model based Arterial Input Function (m-AIF) and accurate individualized AIF (i-AIF) from the femoral artery by comparing their performance in areas suspicious for prostate cancer on endorectal prostate MR at 3.0T. Mean prostate Ktrans values obtained were significantly different using m-AIF and i-AIF. mAIF results in less variable PK analysis results. Further studies are required to determine if mAIF will allow for more robust comparisons in longitudinal studies, or whether mAIF is in fact under-representing underlying areas of tumor

Jarrett Grover¹, and Yousef Mazaheri^2
1Memorial Sloan Kettering Cancer Center, New York, NY, United States, ²Medical Physics, Memorial Sloan Kettering Cancer Center, New York, New York, United States

The present investigation aims to address how physiological properties of solid tumor impact the uptake and distribution of different macromolecular agents used in in vivo anti-tumor therapies in solid tumors using DCE-MRI analysis in human prostate.

Marnix Christiaan Maas¹, Jurgen J. Fütterer¹, and Tom W.J. Scheenen^1
1Department of Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands

High b-value (b>1000 s/mm2) DWI images potentially improve diagnostic accuracy, but are sensitive to artifacts and require long acquisition times for sufficient SNR. In this work we investigated high b-value images obtained by extrapolation from a set of lower b-value images, which aim to combine the diagnostic performance of high b-values with the improved SNR and reduced artifacts of lower b-values. The contrast-to-noise ratio (CNR) between suspicious lesions and normal-appearing tissue in extrapolated b=1400 s/mm2 images was comparable to that in measured images, and results suggested that a slight improvement in CNR may even be achievable with extrapolated images.

Yousef Mazaheri¹, Alvarez Vargas², Oguz Akin², Debra Goldman², and Hedvig Hricak^2
1Medical Physics, Memorial Sloan Kettering Cancer Center, New York, New York, United States, ²Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, United States

In this study, we have compared the ability of mono- and non-monoexponential parameters measured with DW-MRI to identify malignant regions in the PZ and compared the findings with whole-mount pathology.

Jing Ren¹, Yi Huan², and Mengqi Wei^2
1Department of Radiology, Xijing Hospital, Fourth Military Medical University, Xi¡¯an, Shaanxi, China, People's Republic of, ²Xijing Hospital, Fourth Military Medical University, China, People's Republic of

To evaluate the efficacy of apparent diffusion coefficient (ADC) maps combined with T2-weighted (T2w) methods in quantitatively assessing the detection of malignant nodules in the prostatic transition zone (TZ) Quantitative MRI biomarkers play very important roles in the differential diagnosis of cancer and BPH in prostatic TZ. In particular, T2w combined with ADC improves the clinical diagnosis of cancer in the prostatic TZ.

Chunmei Li¹, Min Chen¹, Saying Li¹, Xuna Zhao², Chen Zhang¹, and Cheng Zhou^1
1Beijing Hospital, Beijing, China, People's Republic of, ²Peking University

Purpose: To investigate the characteristics of combined DTI and quantitative DCE-MRI at 3.0 T in differentiating prostate cancer from noncancerous prostatic tissue in peripheral zone. Methods: DTI and DCE-MRI of patients with prostatic diseases were obtained. Parameters of DTI and DCE-MRI were calculated. Receiver operator characteristic curves were generated. Results: There were significant differences in all the parameters between prostate cancer and noncancerous prostatic tissue. The AUC for DTI and DCE-MRI combination was greater than either DTI or DCE-MRI alone. Conclusion: The combination of DTI and DCE-MRI has better accuracy in the diagnosis of prostate cancer than either technique alone.

Valeria Panebianco¹, Sabina Prato ², Daniele Lisi¹, Valeria Buonocore¹, Tommaso Biondi¹, and Roberto Passariello^1
1Department of Radiological Sciences, Sapienza University, Rome, Italy, ²MR Advanced Applications, GE Healthcare, Milan, Italy

In patients submitted to a nerve-sparing prostatectomy is important the depiction of the neuroanatomical distribution and relationship of periprostatic nerve with capsular profile using Diffusion Tensor Imaging (DTI) and tracing the spatial orientation of the neuro component depending of the diffusion process. Fibertracking with endorectal coil of the periprostatic nerve is a feasible application, obtaining a precise neuroanatomical evaluation of the periprostatic nerve in terms of: distance from nerve fiber to prostate capsular profile, distance from nerve fiber to the lesion and integrity and course of the nerve on each part of the prostate.

Liang Wang¹, Hedvig Hricak², and Oguz Akin^2
1Tongji University Hositla of HUST, Wuhan, Hubei, China, People's Republic of, ²Memorial Sloan-Kettering Cancer Center

Synopsis: To evaluate inter-observer variability in assessment of with/without endorectal MRI prostate volumetry as a surrogate for transrectal ultrasound volumetry in estimating iodine – 125 seeds in brachytherapy.

Cynthia Menard^1,2, Douglas Iupati¹, Jenny Lee¹, Anna Simeonov¹, Jessy Abed¹, Julia Publicover¹, Peter Chung¹, Andrew Bayley¹, Charles Catton¹, Michael Milosevic¹, Robert Bristow¹, Gerard Morton³, Padraig Warde¹, Kristy Brock¹, and Masoom Haider^3
1Princess Margaret Hospital, Toronto, ON, Canada, ²Department of Radiation Oncology, University of Toronto, Toronto, Ontario, Canada, ³Odette Cancer Center

We evaluated the performance of MRI (plus/minus guided biopsy) in delineating tumor boundaries for focal salvage therapy of prostate cancer recurrence after external beam radiotherapy. Twenty-three patients were enrolled and received an integrated multiparametric diagnostic and interventional biopsy procedure. MRI alone was not sufficiently accurate to define boundaries for tumor-targeted salvage even with addition of an uncertainty margin. MRI-guided biopsy improved both detection and delineation accuracy for recurrent tumor regions, and changed salvage therapy planning in the majority of patients.

Dorota Jakubowski Wisner¹, Vibhas S. Deshpande², Bonnie N. Joe¹, David A. Porter³, C. Belinda Chang¹, Gerhard A. Laub², and Nola Hylton^1
1Radiology and Biomedical Imaging, Univeristy of California, San Francisco, San Francisco, CA, United States, ²MR Research and Development, Siemens Medical Solutions USA, Inc., San Francisco, CA, United States, ³MR Research and Development, Siemens Medical Solutions, Erlangen, Bavaria, Germany

Diffusion weighted imaging (DWI) has shown promise for improving the specificity of breast MRI, but is hampered by poor resolution, field homogeneity, and overall image quality at 3T. We compare a novel DWI technique named RESOLVE against standard single short spin echo EPI for lesion conspicuity and level of detail in 15 patients, and for ADC values compared to pathology in a smaller sub-selection of these patients. Preliminary results suggest that RESOLVE provides robust, high-resolution DWI at 3T, which may eventually advance the potential clinical utility of DWI in breast imaging.

Shotaro Kanao¹, Tomohisa Okada¹, Mami Iima¹, Kazuna Takeda¹, Shigeaki Umeoka¹, Takeshi Kubo¹, and Kaori Togashi^1
1Diagnostic Imaging and Nuclear Medicine, Kyoto Univerisity Graduate School of Medicine, Kyoto, Kyoto, Japan

Diffusion weighted imaging (DWI) is expected to increase diagnostic accuracy of detecting breast cancer. But compared to dynamic MRI, the imaging quality of DWI is poor because of distortion, poor fat suppression, low spatial resolution and so on. Recently DWI using readout segmented echo planner imaging (RS-EPI) is introduced to our institution as work-in-progress sequence and expected to improve imaging quality. We evaluated the imaging quality of 3.0 T diffusion weighted breast MRI using RS-EPI compared to single shot EPI (SS-EPI). Breast diffusion weighted MRI using RS-EPI can reduce imaging distortion and provide better visualization around the nipple.

Lisa Singer¹, Lisa J Wilmes¹, Emine U. Saritas^2,3, Ajit Shankaranarayanan⁴, Evelyn Proctor¹, Dorota Wisner¹, Belinda Chang¹, Bonnie N. Joe¹, Dwight G. Nishimura³, and Nola M. Hylton^1
1Radiology and Biomedical Imaging, UCSF, San Francisco, CA, United States, ²Department of Bioengineering, UC Berkeley, Berkeley, CA, United States, ³Department of Electrical Engineering, Stanford University, Stanford, CA, United States, ⁴Applied Science Laboratory, GE Healthcare, Menlo Park, CA, United States

Current diffusion-weighted MRI (DW-MRI) sequences used in the breast are often limited in spatial resolution. A reduced field-of-view (rFOV) DW-MRI sequence providing high in-plane resolution was optimized for imaging breast tumors and evaluated against a standard FOV DW-MRI sequence in patients with invasive breast cancer. Quantitative analysis found similar mean tumor apparent diffusion coefficient (ADC) values for both sequences; however, significant differences were found in parameters related to the tumor ADC distribution. Qualitative analysis suggested that tumor depiction on rFOV was improved. Larger studies comparing the ability of rFOV and standard FOV parameters to predict clinical outcomes are needed.

Vibhas S Deshpande¹, Dorota J Wisner², John W Grinstead¹, Thorsten Feiweier³, Bonnie N Joe², and Gerhard A Laub^1
1Siemens Medical Solutions USA, Inc., San Francisco, CA, United States, ²Dept. of Radiology and Biomedical Engineering, UCSF, San Francisco, CA, United States, ³Siemens Medical Solutions, Erlangen, Germany

Breast MRI has high sensitivity in detecting malignancies, but suffers from moderate specificity for lesion characterization. Diffusion imaging has shown promise for better lesion characterization. Diffusion imaging at 3T offers higher SNR, which can translate to higher spatial resolution and improved diagnostic utility. However, single-shot EPI (ss-EPI) diffusion imaging is sensitive to B0 inhomogeneities, which can cause image artifacts. Moreover, in wide-bore (70 cm) 3T scanners, with the higher likelihood of larger patients, the problems of non-uniform B0 are exacerbated. In this work, we demonstrate the efficacy of technical improvements in ss-EPI breast diffusion imaging for improving image quality. These improvements are applicable to all field strengths and scanner types.

David John Manton¹, Filip Van Kove¹, Martin D Pickles¹, and Lindsay W Turnbull^1
1Yorkshire Cancer Research Centre for MR Investigations, Hull-York Medical School, Hull, East Yorkshire, United Kingdom

The accuracy, compared to pathology, of contrast-enhanced MRI-derived, pre-operative maximum tumour diameters was assessed across a range of clinically relevant sub-sets (e.g. tumour type and Her2 status) in a retrospective cohort of women who had undergone neoadjuvant chemotherapy for locally advanced breast cancer. Bland-Altman limits of agreement (LOAs) were not statistically significant within sub-sets so the data-set was analysed as a whole. MRI demonstrated a low false complete response rate (13%), a small bias (-4.85 mm on average) but quite large LOAs (95% of the differences lying between 38.1 and -47.8 mm) thus lending more credibility to its clinical utility.

Li Guo¹, Xiao-ying Wang¹, Nai-shan Qin¹, and Xue-xiang Jiang^1
1Radiology, Peking University First Hospital, Beijing, China, People's Republic of

48 patients underwent MR before and after 4 cycles neoadjuvant chemotherapy(NAC). The ADC value and the maximum diameter of the cancer foci were measured , and the rate of their changes ¨SADC% and ¨SD% were calculated. All the foci were divided into three groups with different Ki-67 index level.One-Way ANOVA was used to determine whether there were differences between the three groups. The result was that the ADC value and the maximal diameter of breast cancer differed with different expression levels of Ki-67 index before and after neoadjuvant chemotherapy, and their response to NAC varied as well.

Mies A. Korteweg¹, Wouter B. Veldhuis¹, Fredy Visser¹, Peter R. Luijten¹, Willem P.Th.M. Mali¹, Paul J. van Diest², Maurice A.A.J. van den Bosch¹, and Dennis W.J. Klomp^1
1Radiology, University Medical Center Utrecht, Utrecht, Netherlands, ²Pathology, University Medical Center Utrecht, Utrecht, Netherlands

Feasibility of 7T breast MRI was evaluated. First the intrinsic sensitivity gain was compared to 3T in healthy volunteers resulting in a 5.7 times increased SNR. Next the clinical application of 7T MRI in breast cancer patients receiving neo-adjuvant chemotherapy was evaluated. High resolution images (450 µm isotropic) were acquired to describe tumor morphology. Apparent diffusion coefficients, tumor size and the total Choline pool were measured during chemotherapy. In one case the 7T results were more consistent with histopathology than 3T. Dedicated 7T breast MRI is technically feasible, can provide more SNR than 3T, and has diagnostic potential.

Hyeon-Man Baek¹, Jeon-Hor Chen², Orhan Nalcioglu², and Min-Ying Su^2
1Advanced Imaging Research Center, UT Southwestern Medical Center, Dallas, TX, United States, ²Tu & Yuen Center for Functional Onco-Imaging, UC Irvine, Irvine, CA, United States

The aim of our study was to determine the feasibility of using quantitative 1H-MRS without water-fat suppression to monitor tumor response to neoadjuvant chemotherapy. Our study showed the reduction in tCho at the first and second follow-up was significantly higher compared with the reduction in the tumor size (mean percentage change -72.6% vs. -9.7%, p < 0.0001; -96.6% vs. -56.7%, p < 0.003). In addition, the reduction in H2O and H2O/CH3 were also significantly higher than the reduction in tumor size at FU-1 (-36.2% vs. -9.7%, p = 0.028; -43.2% vs. -9.7%, p = 0.033). Therefore, the present study demonstrates that in vivo quantitative 1H-MRS without water–fat suppression can be useful for the detection and therapy response monitoring of breast cancer.

April M. Chow¹, Polly S.Y. Cheung², Raymond Lee³, Ka Man Chan³, Sau Fan Liu¹, Siu Ki Yu¹, and Gladys G. Lo^3
1Medical Physics & Research Department, Hong Kong Sanatorium & Hospital, Happy Valley, Hong Kong SAR, China, People's Republic of, ²Breast Care Center, Hong Kong Sanatorium & Hospital, Happy Valley, Hong Kong SAR, China, People's Republic of, ³Department of Diagnostic and Interventional Radiology, Hong Kong Sanatorium & Hospital, Happy Valley, Hong Kong SAR, China, People's Republic of

Diffusion-weighted imaging (DWI) has been widely used to characterize malignant and benign lesions at 1.5 T. However, conventional DWI uses 5 - 8 mm slice thickness, partial volume averaging for small lesion and may lead to difficulty in distinguishing it from neighboring normal tissue. The use of thinner slices may allow small, low-contrast lesions to be detectable, improving the depiction of small tumors. In this study, we evaluated the effect of thin-section DWI in detecting malignant breast lesions at 3 T. Reduced ADC values were observed and characterized in malignant lesions, compared to normal breast tissues. Differential ADC values were demonstrated between conventional and thin-section DWI in malignant lesions. Sensitivity increased from 80% to 98% with thin-section DWI in diagnosis of malignant lesions, as compared to conventional DWI.

Giuseppe Petralia¹, Luke Bonello², Paul Summers¹, Lorenzo Preda¹, Roberto Di Filippi¹, Moreno Pasin¹, Marzia Locatelli³, Giuseppe Curigliano³, and Massimo Bellomi^1,2
1Radiology, European Institute of Oncology, Milan, Milan, Italy, ²School of Radiology, University of Milan, Milan, Italy, ³Medical Oncology, European Institute of Oncology, Milan, Milan, Italy

The clinical course of breast cancer (BCa) depends on several molecular and histological characteristics. The aim of this study was to correlate apparent diffusion coefficient (ADC) with molecular and histological prognostic factors of BCa. Fifty-three BCa patients underwent DW-MRI of the breast. ADC was lower for HER-2 than for triple receptor negative subtype, and for T3 than for T1 BCa (P=.055 and .03, respectively). There was a marginally significant correlation (Spearman r=-0.35, P<.05) between ADC and HER-2 expression. There was no correlation between ADC and variables examined, with possible exception for HER-2. T3 subgroup may bear a distinct ADC behaviour.

Elizabeth Anne Maxine O'Flynn¹, Marco Borri¹, Maria Schmidt¹, Veronica Morgan¹, Sharon Giles¹, Catherine Parry-Jones¹, and Nandita M deSouza^1
1Clinical Magnetic Resonance, Cancer Research UK and EPSRC Cancer Imaging Centre, Sutton, Surrey, United Kingdom

ADC values during the secretory phase (days 15-28) of the menstrual cycle are known to be slightly higher and less restricted due to the increased water content within the breast. We have demonstrated significant differences in mean ADC values relating to menopausal state, with the ADC value being higher in premenopausal volunteers in the secretory phase (days 15-28) compared to postmenopausal volunteers (p=0.008) due to the effects of increased serum oestradiol and progesterone. This can have implications in the detection of tumours and increased false positives in dynamic contrast-enhanced breast MRI due to the high background parenchymal enhancement.

Liuquan Cheng¹, Yuhan Bai^1,2, Jing Zhang^1,3, Mei Liu⁴, and Xiru Li^5
1Radiology, Chinese PLA General Hospital, Beijing, Beijing, China, People's Republic of, ²Radiology, The People's Hospital of Wuhan University, Hubei, ³Radiology, Chinese PLA Navy General Hospital, Beijing, ⁴Pathology, Chinese PLA General Hospital, Beijing, Beijing, China, People's Republic of, ⁵Surgery, Chinese PLA General Hospital, Beijing, Beijing, China, People's Republic of

We found that the cut-off point ADC value for differential diagnosis differed between mass and non-mass-like enhancement lesions and should be weighted in malignancy prediction. The optimized cutoff point ADC value for mass lesion is 1.05¡Á10-3mm2/s, which is lower than the NMLE where ADC=1.35¡Á10-3mm2/s and different than the benign and malignant without MRI types differentiation (ADC=1.25¡Á10-3mm2/s). This finding was different than the others studies whose ADC measurement was not affected by the lesion size.

Christian Geppert¹, Matthias Fenchel¹, Rolf Janka², Andre de Oliveira¹, Berthold Kiefer¹, Michael Uder², and Evelyn Wenkel^2
1Siemens Healthcare, Erlangen, Germany, ²Radiologisches Institut, Universitätsklinikum Erlangen, Erlangen, Germany

In diagnostic breast imaging, usually a dynamic scan is performed which consists of a native scan and a number (usually 3-10) of post contrast-administration (CA) scans of the identical volume. Despite the use of automatic injectors, the timing of the injection with the scan protocol is subject to operator dependent variations such as pause durations, patient weight- and injection location dependent changes etc. Moreover, due to the widespread use of CAD workstations that work with a fixed threshold for a defined time point such as 50 or 100% at the first post-CA time point, timing variations can significantly influence the appearance of enhancement maps. The goal of this work was to implement a bolus detection method for breast MRI in a routine clinical situation that eliminates many patient and operator dependent variations, helps to avoid re-scans due to CA problems and thus might improve the accuracy of dynamic breast protocols.

Kyunghyun Sung¹, Bruce L Daniel¹, and Brian A Hargreaves^1
1Radiology, Stanford University, Stanford, California, United States

Variable flip angle spoiled gradient echo (SPGR) acquisitions, called DESPOT1, are a common choice to measure T1 since they can provide a fast 3D volumetric T1 mapping. DESPOT1, however, heavily depends on the set of flip angles used, and therefore is sensitive to any flip angle variation. Transmit B1 field (B1) inhomogeneity creates flip angle variation and the variation tends to be 30 - 50% across the breast at 3T. In this work, we include B1 mapping in our breast DCE imaging protocol, and compensate T1 maps by including B1 variation in the DESPOT1 calculation. We then compare T1 relaxation in fat (as a validation) with and without compensating B1 variation in a total of 25 patients at 3T.

Manojkumar Saranathan¹, Brian A Hargreaves¹, Catherine J Moran¹, and Bruce Daniel^1
1Radiology, Stanford University, Stanford, CA, United States

Dynamic-contrast enhanced (DCE) MRI is a standard element of clinical breast protocols, and is challenging due to the need for both high spatial and high temporal resolution. High temporal resolution is needed for accurate pharmacokinetic modeling and high spatial resolution for tumor characterization. Although some imaging protocols include a combination of rapid, low-resolution scans and slower high-resolution scans, these can cause practical complications such as delays and/or accidental changes to parameters (like gain) that could render the acquisition useless for quantifying contrast uptake. Here we present a simple, single scan technique that acquires images of variable temporal and spatial resolution, with the timing tailored to the contrast dynamics in breast imaging.

Bonnie N. Joe¹, Dorota Wisner¹, Vignesh A Arasu¹, Sachiko Suzuki¹, Vibhas S. Deshpande², Belinda Chang¹, Gerhard Laub², and Nola M Hylton^1
1Dept of Radiology and Biomedical Imaging, UCSF, San Francisco, CA, United States, ²Siemens Medical Solutions USA, Inc, San Francisco, CA, United States

We evaluated lesion conspicuity and level of detail on breast MR images of patients scanned with radial-centric and linear ordered k-space techniques on separate days. Breast imagers ranked linear ordering better than radial-centric for lesion conspicuity and level of image detail. Similarly, quantitative analyses showed increased peak signal enhancement ratio (SER) in 4/5 patients and increased SER total enhancing volume in 5/5 patients for linear versus radial-centric. More motion artifact was noted with linear ordering but overall image quality ranked similarly between linear and radial-centric. Radiologists should be aware of these tradeoffs when designing their breast MR cancer imaging protocol.