ORIGINAL ARTICLE

Apparent Diffusion Coefficient Value Is Not Dependent on Magnetic Resonance Systems and Field Strength Under Fixed Imaging Parameters in Brain Akio Ogura, PhD,* Takayuki Tamura, PhD,† Masanori Ozaki, MS,‡ Tsukasa Doi, RT,§ Koji Fujimoto, MD, PhD,|| Tosiaki Miyati, PhD, DMSc,¶ Yukiko Ito, MS,# Fumie Maeda, RT,** Hiroyuki Tarewaki, RT,§ and Mitsuyuki Takahashi, RT††

Objective: The aim of the study was to investigate the causes of apparent diffusion coefficient (ADC) measurement errors and to determine the optimal scanning parameters that are independent of the field strength and vendors of the magnetic resonance (MR) system. Materials and Methods: Brain MR images of 10 healthy volunteers were scanned using 6 MR scanners of different field strengths and vendors in 2 different institutions. Ethical review board approvals were obtained for this study, and all volunteers gave their informed consents. Coefficient of variation (CV) of ADC values were compared for their differences in various MR scanners and in the scanned subjects. Results: The CV of ADC values for 6 different scanners of 6 brains was 3.32%. The CV for repeated measurements in 1 day (10 scans per day) and in 10 days (scan per day for 10 days) for 1 subject was 1.72% and 2.96%, respectively (n = 5, P < 0.001). The CV of measurements for 10 healthy subjects was 5.22%. The measurement errors of the ADC values for 6 different MR units in 1 subject were higher than the intrascanner variance for the same subject but were lower than the intersubject variance for the same scanner. Conclusions: The variance in the ADC values for different MR scanners is reasonably small if appropriate scanning parameters (repetition time, >3000 ms; echo time, minimum; and high enough signal-to-noise ratio of high-b diffusion-weighted image) are used. Key Words: apparent diffusion coefficient, signal-to-noise ratio, b value, coefficient of variation of ADC, different field strength (J Comput Assist Tomogr 2015;39: 760–765)

T

he signal intensity of T1- and T2-weighted images or the vascularity of contrast enhancement has been used conventionally to differentiate between benign and malignant lesions and also to aid in predicting therapeutic efficacy on magnetic resonance (MR) imaging. However, one may want to avoid the use of contrast media because of its cost and because of safety concerns about the reported adverse effect of nephrogenic systemic fibrosis. From the *Graduate School, Gunma Prefectural College of Health Sciences, Maebashi; †Department of Clinical Support, Hiroshima University Hospital, Hiroshima; ‡Division of Health Sciences, Kanazawa University Graduate School of Medical Sciences, Kanazawa; §Department of Medical Technology, Osaka University Hospital, Osaka; ||Department of Diagnostic Imaging and Nuclear Medicine, Graduate School of Medicine, Kyoto University, Kyoto; ¶Institute of Medical, Pharmaceutical and Health Sciences, Kanazawa University, Kanazawa; #Japan Community Health Care Organization Sendai Minami Hospital, Sendai; **Department of Radiology, Kyoto City Hospital, Kyoto; and ††Department of Radiology, Yokohama Sakae Kyosai Hospital, Federation of National Public Service Personnel Mutual Associations, Yokohama, Japan. Received for publication January 20, 2015; accepted March 16, 2015. Correspondence to: Akio Ogura, PhD, Graduate School, Gunma Prefectural College of Health Sciences, 323-1, Kamioki-machi, Maebashi, Gunma, Japan (e‐mail: [email protected]). The authors declare no conflict of interest. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. DOI: 10.1097/RCT.0000000000000266

760

www.jcat.org

A particular advantage of diffusion-weighted MR imaging (DWMRI) is that it does not require intravenous contrast media. Diffusion-weighted MR imaging originally was an important tool in the diagnosis of acute stroke of the brain. Subsequently, whole-body DW-MRI scanning was developed along with parallel imaging and is now being employed for the detection of tumors. Currently, it can be used for differentiation of tumors by the ADC values calculated from DW-MRI. Its clinical use now includes improved tissue characterization for monitoring treatment response after chemotherapy or radiation and for differentiating posttherapeutic changes from residual active tumor. Many studies have been conducted using ADC to distinguish malignant tumors from benign tumors, particularly in pharmaceutical drug development and also for predicting therapeutic efficacy in MR imaging.1–31 However, a significant discrepancy exists between the threshold values of ADC as used by authors of different articles, although these studies evaluated the same target region. These reports are dependent on the specific institutions in which the studies were performed and the devices employed in those studies. Therefore, at present, the ADC cannot be used as a quantitative value.32,33 To ensure consistent evaluation and interpretation of quantitative ADC values obtained using different vendors' units and in different institutions, it is absolutely necessary to ensure agreements among all stakeholders on standards for acquisition protocols, repeatability/reproducibility, and postprocessing procedures. The diffusion coefficient values should theoretically be the same in all organizations and should not depend on field strength and vendors. The purpose of our study was to find the causes of such measurement errors, recommend the optimal scanning parameters, and show that under a set of optimal scanning parameters, ADC values are independent of differences in the system field strengths and vendor equipment.

MATERIALS AND METHODS Scanning Parameters Eddy currents related to the diffusion gradient and echo planar imaging techniques lead to geometric distortions, which can be reduced by increasing readout bandwidths or reducing the echo spacing, conversely reducing the length of the image readout. Therefore, parallel imaging and a rectangular field of view were used to reduce image distortion for clinical imaging. We used the parallel imaging factor of 2 and the rectangular field of view of 80% for clinical evaluations. The minimum short echo time (TE) and repitition time (TR) greater than 3000 milliseconds were used to zavoid the influence of T2 relaxation and T1 relaxation in

J Comput Assist Tomogr • Volume 39, Number 5, September/October 2015

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

J Comput Assist Tomogr • Volume 39, Number 5, September/October 2015

ADC Values Are Not Dependent on the System Units

TABLE 1. The Specifications of the 6 MR Scanners in the Present Study Vender GE Healthcare Philips Healthcare Siemens Medical Solutions GE Healthcare Siemens Medical Solutions Toshiba Medical Systems

Institute

Field Strength, T

Maximum Gradient Strength, mT/m

Slew Rate, T/m per s

Osaka University Osaka University Kyoto University Osaka University Kyoto University Kyoto University

3 3 3 1.5 1.5 1.5

40 40 45 30 45 33

150 200 200 120 200 200

findings obtained by different institutions.32,34 For calculation of the ADC value, the 2-point method (later) is applied in this study. ADC value ¼ In ½SIðb1Þ = SIðb2Þ  = ðb1− b2Þ

ð1Þ

Effect of Signal-to-Noise Ratio on ADC Measurements The inaccurate ADC values calculated for different scanning parameters affecting the signal-to-noise ratio (SNR) were measured using detergent phantoms. The detergent phantoms

(surface-active agent, 18%; T1 value, 620 ms; T2 value, 321 ms) were scanned with different NEX (number of excitations), slice thickness, field of view (FOV), and matrix size. Basic scanning parameters were as follows: TR, 10,000 ms; TE, 138 ms; slice thickness, 5 mm; FOV, 300 mm; and matrix size, 128  128. The other parameters were fixed, and the phantom was scanned with the conditions that the FOV was changed between 212  212, 300  300, and 424  424 mm. The slice thickness was changed between 2, 5, and 10 mm. The NEX was changed between 1, 4, and 16, and the matrix size was changed between 64, 128, and 256. The b values were varied between 0

FIGURE 1. A, ADC values for different NEX. There was no change in ADC values for different NEX. The increase of NEX did not influence the precision of the ADC value in a statistically significant manner (P < 0.005). B, ADC values for different slice thickness. The precision of the ADC value was maintained when a high b value with a high SNR and a thick slice was used. C, ADC values for different FOV. The precision of the ADC value was maintained when a high b value with high SNR and large FOV was used. D, ADC values for different matrix sizes. The precision of the ADC value was maintained for high b values with high SNR and reduced numbers of matrix size. © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

www.jcat.org

761

J Comput Assist Tomogr • Volume 39, Number 5, September/October 2015

Ogura et al

and 8000 s/mm2, where b is 0, 500, 1000, 2000, 2500, 3000, 3500, 4000, 5000, 6000, 7000, and 8000 s/mm2. The ADC was calculated using the 2-point of the previous b values. The device used was a 1.5T MR system (SIGNA HDxt; GE Healthcare, Milwaukee, Wis).

Dependency of ADC Values on Scanning Parameters in Clinical Evaluation

RESULTS Effect of SNR on ADC Measurements The ADC values for a varying NEX, slice thickness, FOV, and matrix size are shown in Figures 1A-D. Generally, the ADC value does not depend on the b value by the free diffusion field.

The 6 human brains were scanned using 6 MR scanners of different field strengths and vendors in 2 different institutions (Osaka University Hospital and Kyoto University Hospital). We prospectively examined each of the 6 healthy volunteers (men, 6; age range, 27–56 years) within 2 days by using each of the 6 MR scanners. The devices used were 3T MR systems by GE Healthcare, Philips Medical Systems (Best, the Netherlands), and Siemens (Erlangen, Germany) and 1.5T MR systems by GE Healthcare, Siemens, and Toshiba (Tokyo, Japan). The specifications of each device are shown in Table 1. In both institutions, ethical review board approvals were obtained for this study and all volunteers gave their informed consent. The scanning parameters were as follows: TR, 2000 and 8000 ms; slice thickness, 2 and 5 mm; NEX, 1, 2, and 4; TE, minimum; and b values, 0 and 1000 s/mm2. Total scanning were 12 times at 1 subject. The scanning procedures were completed in 2 days to reduce the influence of time. Regions of interest of 100 mm2 was set on the same point on the left frontal lobe white matter on a slice of the basal nucleus. The ADCs were calculated with the mean of 3 sets of regions of interest using a locally developed code for the Matlab2011a application software. Each ADC value was statistically analyzed using Kruskal-Wallis and Tukey tests. These analyses were performed using StatMate Software for Windows, Version 5.

Relationship of ADC Values to b Values on Clinical Evaluation As the second study, the 10 healthy subjects (men, 8; women, 2; aged, 27–56 years; mean age, 35.6 years) were scanned within 2 days using the 6 MR scanners of varying field strengths and vendors in the 2 institutions. The devices used were the same as the previous. The scanning parameters were fixed with highly precise measurements of ADC in this study (TR, 3000 ms; TE, minimum; slice thickness, 5 mm; NEX, 2). The b values used were 0, 100, 500, 1000, or 2000 s/mm2. The scanning procedures were completed in 2 days. The ADCs of each device and each subject were calculated from b values of 0 to 500, 0 to 1000, 0 to 2000, 100 to 1000, and 500 to 2000 s/mm2. The averaged coefficient of variation (CV) value of ADC among the 6 scanners for each subject was then calculated. The CV was calculated by dividing the SD by average with 6 devices.

Comparison of the Measurement Error of ADC Values With Multiple Scanning One healthy subject was scanned 10 times in 1 day and then 10 times once per day for 10 days, with b values of 0 to 2000 s/mm2 using 3T MR systems by GE Healthcare. The ADC values were calculated with b values of 0 and 1000 s/mm2. The CV values between the 10 scans (10 times in 1 day and 10 times once per day for 10 days) for this individual were evaluated. In addition, the averaged CV values between the 10 healthy subjects for the same scanner were calculated. The CV value of the ADC between the 6 different scanners was compared with the other variable factors.

762

www.jcat.org

FIGURE 2. A, ADC values for 1, 2, and 4 NEX at 1 person scanned on 6 MR systems. The NEX did not affect the ADC value in a statistically significant manner (P < 0.005). B, ADC values with TR of 2000 and 8000 milliseconds at 1 person scanned on 6 MR systems. The ADC value in TR of 2000 milliseconds varied slightly. C, ADC values with slice thickness of 2 and 8 mm at 1 person scanned on 6 MR systems. The ADC values at 2-mm thickness were higher and varied. © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

J Comput Assist Tomogr • Volume 39, Number 5, September/October 2015

ADC Values Are Not Dependent on the System Units

Therefore, we defined the ADC of the low b value as true value and defined the difference in the ADC with the high b values as error of measurement. Precise ADC value for high b values was obtained by high SNR, thick slices, large FOV, and small matrix size. The error of ADC values was increased when SNR was low. However, for NEX changed, there was no change in the ADC value. An increase in NEX did not influence the error of the ADC value in a statistically significant manner (P < 0.005).

Dependency of ADC Values on Scanning Parameters on Clinical Evaluation The ADC values for NEX of 1, 2, and 4 at 1 subject scanned on 6 MR systems are shown in Figure 2A. The NEX did not have a statistically significant effect on the ADC value (P < 0.005). The ADC values for TR of 2000 and 8000 milliseconds at 1 subject scanned on 6 MR systems are shown in Figure 2B. The ADC value when TR is 2000 milliseconds varied slightly. The ADC values for slice thickness of 2 mm and 5 mm at 1 subject scanned on 6 MR systems are shown in Figure 2C. The ADC value at 2-mm thickness was higher and variable.

Dependency of ADC Values on Different b Values in Clinical Evaluation Figure 3 indicated the ADC values calculated from different b values scanned by 6 MR systems at 1 subject. The ADC values were calculated at b values of 0 to 500, 0 to 1000, 0 to 2000, 100 to 1000, and 500 to 2000 s/mm2. The ADC values while using different devices with different magnetic field strengths did not vary for each person. However, the b values used for the ADC calculation influenced the ADC value. The means and CV of ADCs (b = 0-1000 s/mm2) for 10 subjects scanned on 6 MR systems are shown in Figure 4. The maximum and mean CVs were 4.65% and 3.6%, respectively. The averaged ADCs calculated from b values of 0 to 500, 0 to 1000, 100 to 1000, 0 to 2000, and 500 to 2000 s/mm2 were shown in Figure 5. The ADC value changed greatly with the b values used for calculation.

FIGURE 3. The ADC values calculated by using different b values at 1 person scanned on 6 MR systems. The ADC values were calculated at b values of 0 to 500, 0 to 1000, 0 to 2000, 100 to 1000, and 500 to 2000 s/mm2. The ADC values from different devices with different magnetic field strengths did not vary.

FIGURE 4. The means and CV of ADCs (b = 0-1000) for 6 scanners in 10 persons. The CVs were 0.0465 as the maximum and 0.036 as the mean.

Comparison of the Measurement Error of ADC Values With Multiple Scans The DW-MRI signals for 1 healthy subject scanned once per day for 10 days with a b value of 0 to 2000 s/mm2 are shown in Figure 6A. The CV values between the 10 scans, namely, 10 times in 1 day and once per day for 10 days for 1 subject were 1.72 and 2.96%, respectively. Figure 6B shows the DW-MRI signals of 1 healthy subject scanned using 6 different scanners with a b value of 0 to 2000 s/mm2. The averaged CV value of the ADC values between the 6 scanners was 3.32%. The CV value between 10 healthy subjects for the same scanner was 5.22%.

DISCUSSION We recognized that the choice of b values greatly affects the calculation of ADC. Secondly, the ADC value was affected by the SNR of high-b DW images. Therefore, it was important to assess whether the SNR of high-b DW images was high enough for a proper ADC calculation. In addition, an increase in the NEX did not increase the SNR of high-b DW image and did not affect the precision of the ADC value. For the reason of this, the averaging method of DWI is to average the frequency domain of k-space after the Fourie transfer as compared with other scan techniques. This is to prevent that the addition on k-space makes diffusive random characteristics averaged. Therefore, the increase of NEX did not improve SNR of high-b DWI and not lead to the precision of ADC.

FIGURE 5. The averaged ADC values were calculated by using different b values. The ADC value changed greatly by the use of different b values for the calculation.

© 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

www.jcat.org

763

Ogura et al

J Comput Assist Tomogr • Volume 39, Number 5, September/October 2015

example, an ADC (50-2000) = *** would mean b values of 50 and 2000 s/mm2. The averaged CV value of ADC among the 6 scanners using optimal scanning parameter was 3.32%. The CV values between scans 10 times in 1 day and 10 times once per day for 10 days for 1 subject were 1.72 and 2.96%, respectively. Moreover, the CV for values between 10 healthy subjects for the same scanner was 5.22%. The measurement errors of ADC values for the 6 devices in 1 subject were higher than those in the ADC values for 10 scans of the same subject in the same device. The measurement errors of the ADC values were however lower when 10 subjects were scanned with the same device. Therefore, even if one uses a scan with a different field strength and from different vendors, the measurement errors of ADC values can be minimized by using optimal scanning parameters. The ADC calculated by same b values cannot theoretically change even if static magnetic field strength and vendors are different. However, the maximum gradient strength and duration of MPG pulse may influence ADC value by vendor's device slightly. However, we believe that the same ADC value can be obtained by any MRI device when scanning parameters are optimized and similar b values are used.

CONCLUSIONS The choice of b values for the calculation greatly affects the ADC value. In addition, ADC value is affected by the SNR of high-b DW images. Therefore, it is important to assess whether the SNR of high-b DW images is high enough for a proper ADC calculation. In addition, an increase in the NEX does not increase the SNR and therefore does not affect the precision of the ADC. The TR greater than 3000 milliseconds should be selected and TE should be set to minimum. With these scanning parameters and enough SNR, which tissue can distinguish in high b value, the ADC value calculated from the same b value is independent of field strength and different vendor machines.

RECOMMENDATIONS FIGURE 6. A, The DW-MRI signal of 1 healthy volunteer scanned 10 times for 10 days with b values of 0 to 2000 s/mm2. The CV value between 10 scans (10 times per 10 days) for 1 person was 2.96%. B, The DW-MRI signal of 1 healthy volunteer using 6 different scanners with b values of 0 to 2000 s/mm2. The DW-MRI signal was shown as high b values on both. The averaged CV value of ADC between the 6 scanners was 3.32%.

The large FOV, thick slices, and small matrix size should be used to increase SNR. In other words, only 1 NEX is enough for the purpose of ADC measurement. The TR greater than 3000 milliseconds should be selected and TE should be set to the minimum value.32 It has been reported that the diffusion coefficient of a human body becomes biexponential.35–38 When perfusion is taken into account, it may be at least triexponential.39,40 Therefore, it should be understood that ADC values vary with selected b values. We think that the reason for the different thresholds of ADC values between malignant and benign tumor noted between different literature reports is related to the difference in b values used for the calculation. Therefore, we recommend that b values used to calculate ADC should be shown with the ADC value. For example, if the ADC was calculated from b values of 0 and 1000 s/mm2, one should indicate ADC (0-1000) = ***. Alternatively, as another

764

www.jcat.org

Apparent diffusion coefficient values vary along with the selected b values for the calculation. Therefore, we recommend that b values used to calculate ADC be clearly shown along with the ADC value. REFERENCES 1. Higano S, Yun X, Kumabe T, et al. Malignant astrocystic tumors: clinical importance of apparent diffusion coefficient in prediction of grade and prognosis. Radiology. 2006;241:839–846. 2. Bulakbasi N, Guvenc I, Onguru O, et al. The added value of the apparent diffusion coefficient calculation to magnetic resonance imaging in the differentiation and grading of malignant brain tumors. J Comput Assist Tomogr. 2004;28:735–746. 3. Muti M, Aprile I, Principi M, et al. Study on the variations of the apparent diffusion coefficient in areas of solid tumor in high grade gliomas. Magn Reson Imaging. 1999;20:635–641. 4. Park MJ, Cha ES, Kang BJ, et al. The role of diffusion-weighted imaging and the apparent diffusion coefficient (ADC) values for breast tumors. Korean J Radiol. 2007;8:390–396. 5. Yerli H, Agildere AM, Aydin E, et al. Value of apparent diffusion coefficient calculation in the differential diagnosis of parotid gland tumors. Acta Radiol. 2007;48:980–987. 6. Matsushima N, Maeda M, Takamura M, et al. Apparent diffusion coefficients of benign and malignant salivary gland tumors. Comparison to histopathological findings. J Neuroradiol. 2007;34:183–189.

© 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

J Comput Assist Tomogr • Volume 39, Number 5, September/October 2015

7. desouza NM, Reinsberg SA, Scurr ED, et al. Magnetic resonance imaging in prostate cancer: the value of apparent diffusion coefficients for identifying malignant nodules. Br J Radiol. 2007;80:90–95. 8. Eida S, Sumi M, Sakihama N, et al. Apparent diffusion coefficient mapping of salivary gland tumors: prediction of the benignancy and malignancy. AJNR AM J Neuroradiol. 2007;28:116–121. 9. Nakayama T, Yoshimitsu K, Irie H, et al. Diffusion-weighted echo-planar MR imaging and ADC mapping in the differential diagnosis of ovarian cystic masses: usefulness of detecting keratinoid substances in mature cystic teratomas. J Magn Reson Imaging. 2005;22:271–278. 10. Guo Y, Cai YQ, Cai ZL, et al. Differentiation of clinically benign and malignant breast lesions using diffusion-weighted imaging. J Magn Reson Imaging. 2002;16:172–178. 11. Sinha S, Lucas-Quesada FA, Sinha U, et al. In vivo diffusion-weighted MRI of the breast: potential for lesion characterization. J Magn Reson Imaging. 2002;15:693–704. 12. Taouli B, Vilgrain V, Dumont E, et al. Evaluation of liver diffusion isotropy and characterization of focal hepatic lesions with two single-shot echo-planar MR imaging sequences: prospective study in 66 patients. Radiology. 2003;226:71–78. 13. Woodhams R, Matsunaga K, Iwabuchi K, et al. Diffusion-weighted imaging of malignant breast tumors: the usefulness of apparent diffusion coefficient (ADC) value and ADC map for the detection of malignant breast tumors and evaluation of cancer extension. J Comput Assist Tomogr. 2005;29:644–649. 14. Pickles MD, Gibbs P, Lowry M, et al. Diffusion changes precede size reduction in neoadjuvant treatment of breast cancer. Magn Reson Imaging. 2006;24:843–847. 15. Yankeelov TE, Lepage M, Chakravarthy A, et al. Integration of quantitative DCE-MRI and ADC mapping to monitor treatment response in human breast cancer: initial results. Magn Reson Imaging. 2007;25:1–13. 16. Theilmann RJ, Borders R, Trouard TP, et al. Changes in water mobility measured by diffusion MRI predict response of metastatic breast cancer to chemotherapy. Neoplasia. 2004;6:831–837. 17. Mardor Y, Pfeffer R, Spiegelmann R, et al. Early detection of response to radiation therapy in patients with brain malignancies using conventional and high b-value diffusion-weighted magnetic resonance imaging. J Clin Oncol. 2003;21:1094–1100. 18. Kul S, Cansu A, Alhan E, et al. Contribution of diffusion-weighted imaging to dynamic contrast-enhanced MRI in the characterization of breast tumors. AJR Am J Roentgenol. 2011;196:210–217. 19. Rechichi G, Galimberti S, Signorelli M, et al. Endometrial cancer: correlation of apparent diffusion coefficient with tumor grade, depth of myometrial invasion, and presence of lymph node metastases. AJR Am J Roentgenol. 2011;197:256–262.

ADC Values Are Not Dependent on the System Units

between squamous cell carcinomas and malignant lymphomas of the head and neck. AJNR Am J Neuroradiol. 2005;26:1186–1192. 24. Sumi M, Ichikawa Y, Nakamura T. Diagnostic ability of apparent diffusion coefficients for lymphomas and carcinomas in the pharynx. Eur Radiol. 2007;17:2631–2637. 25. Zhang J, Tehrani YM, Wang L, et al. Renal masses: characterization with diffusion-weighed MR imaging–a preliminary experience. Radiology. 2008;247:458–464. 26. Yoshikawa T, Kawamitsu H, Mitchell DG, et al. ADC measurement of abdominal organs and lesions using parallel imaging technique. AJR Am J Roentgenol. 2006;187:1521–1530. 27. Maeda M, Matsumine A, Kato h, et al. Soft-tissue tumors evaluated by line-scan diffusion-weighted imaging: influence of myxoid matrix on the apparent diffusion coefficient. J Magn Reson Imaging. 2007;25: 1199–1204. 28. Hambrock T, Somford DM, Huisman HJ, et al. Relationship between apparent diffusion coefficients at 3.0-T MR imaging and Gleason grade in peripheral zone prostate cancer. Radiology. 2011;259:453–461. 29. Oto A, Yang C, Kayhan A, et al. Diffusion-weighted and dynamic contrast-enhanced MRI of prostate cancer: correlation of quantitative MR parameters with Gleason score and tumor angiogenesis. AJR Am J Roentgenol. 2011;197:1382–1390. 30. Peng Y, Jiang Y, Yang C, et al. Quantitative analysis of multiparametric prostate MR images: differentiation between prostate cancer and normal tissue and correlation with Gleason score–a computer-aided diagnosis development study. Radiology. 2013;267:787–796. 31. Ei Khouli RH, Jacobs MA, Mezban SD, et al. Diffusion-weighted imaging improves the diagnostic accuracy of conventional 3.0-T breast MR imaging. Radiology. 2010;256:64–73. 32. Ogura A, Hayakawa K, Miyati T, et al. Imaging parameter effects apparent diffusion coefficient determination of magnetic resonance imaging. Eur J Radiol. 2011;77:185–188. 33. Jambor I, Merisaari H, Aronen HJ, et al. Optimization of b-value distribution for biexponential diffusion-weighted MR imaging of normal prostate. J Magn Reson Imaging. 2014;39:1213–1222. 34. Le Bihan D, Poupon C, Amadon A, et al. Artifacts and pitfalls in diffusion MRI. J Magn Reson Imaging. 2006;24:478–488. 35. Schwarcz A, Bogner P, Meric P, et al. The existence of biexponential signal decay in magnetic resonance diffusion-weighted imaging appears to be independent of compartmentalization. Magn Reson Med. 2004;51: 278–285. 36. Niendorf T, Dijkhuizen RM, Norris DG, et al. Biexponential diffusion attenuation in various states of brain tissues: implications for diffusion-weighted imaging. Magn Reson Med. 1996;36:847–857.

20. Khalid L, Carone M, Dumrongpisutikul N, et al. Imaging characteristics of oligodendrogliomas that predict grade. AJR Am J Roentgenol. 2012;33: 852–857.

37. Mulkern RV, Vajapeyam S, Robertson RL, et al. Biexponential apparent diffusion coefficient parameterization in adult vs newborn brain. Magn Reson Imaging. 2001;19:659–668.

21. Sandrasegaran K, Patel AA, Ramaswamy R, et al. Characterization of adrenal masses with diffusion-weighted imaging. AJR Am J Roentgenol. 2011;197:132–138.

38. Chin CL, Wehrli FW, Hwang SN, et al. Biexponential diffusion attenuation in the rat spinal cord: computer simulations based on anatomic images of axonal architecture. Magn Reson Med. 2002;47:455–460.

22. Razek AA, Sadek AG, Kombar OR, et al. Role of apparent diffusion coefficient value in differentiation between malignant and benign solitary thyroid nodule. AJNR Am J Neuroradiol. 2008;29:563–568.

39. Mulkern RV, Gudbjartsson H, Westin CF, et al. Multi-component apparent diffusion coefficients in human brain. NMR Biomed. 1999;12:51–62.

23. Maeda M, Kato H, Sakuma H, et al. Usefulness of the apparent diffusion coefficient in line scan diffusion-weighted imaging for distinguishing

40. Hayashi T, Miyati T, Takahashi J, et al. Diffusion analysis with triexponential function in liver cirrhosis. J Magn Reson Imaging. 2013;38: 148–153.

© 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

www.jcat.org

765

Apparent Diffusion Coefficient Value Is Not Dependent on Magnetic Resonance Systems and Field Strength Under Fixed Imaging Parameters in Brain.

The aim of the study was to investigate the causes of apparent diffusion coefficient (ADC) measurement errors and to determine the optimal scanning pa...
555KB Sizes 1 Downloads 7 Views