Retrospective Study
Copyright ©The Author(s) 2017. Published by Baishideng Publishing Group Inc. All rights reserved.
World J Hepatol. May 18, 2017; 9(14): 657-666
Published online May 18, 2017. doi: 10.4254/wjh.v9.i14.657
Image quality and diagnostic performance of free-breathing diffusion-weighted imaging for hepatocellular carcinoma
Yukihisa Takayama, Akihiro Nishie, Yoshiki Asayama, Kousei Ishigami, Daisuke Kakihara, Yasuhiro Ushijima, Nobuhiro Fujita, Ken Shirabe, Atsushi Takemura, Hiroshi Honda
Yukihisa Takayama, Department of Radiology, Kitakyushu Municipal Medical Center, Kitakyushu, Fukuoka 802-0077, Japan
Akihiro Nishie, Yoshiki Asayama, Kousei Ishigami, Daisuke Kakihara, Yasuhiro Ushijima, Nobuhiro Fujita, Hiroshi Honda, Department of Clinical Radiology, Graduate School of Medical Sciences, Kyushu University, Fukuoka 812-8582, Japan
Ken Shirabe, Department of Hepatobiliary and Pancreatic Surgery, Gunma University, Graduate School of Medicine, Maebashi, Gunma 371-8511, Japan
Atsushi Takemura, Philips Electronics Japan, Minato-ku, Tokyo 108-8507, Japan
Author contributions: Takayama Y designed and performed the research and wrote the paper; Nishie A designed the research and supervised the report; Asayama Y, Ishigami K, Kakihara D, Ushijima Y and Fujita N designed the research and contributed to the analysis; Shirabe K and Takemura A provided clinical and technical advice; Honda H supervised the report.
Institutional review board statement: This study was reviewed and approved by the Ethics Committee of the Kyushu University Hospital.
Informed consent statement: This study was approved by the Institutional Review Board of our institute. The requirement for written informed consent was waived due to the retrospective nature of the study. For full disclosure, the details of the study are published on the home page of Kyushu University.
Conflict-of-interest statement: We have no financial relationships to disclose.
Data sharing statement: No additional data are available.
Open-Access: This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Correspondence to: Akihiro Nishie, MD, PhD, Associated Professor, Department of Clinical Radiology, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan. anishie@radiol.med.kyushu-u.ac.jp
Telephone: +81-92-6425695 Fax: +81-92-6425708
Received: November 16, 2016
Peer-review started: November 17, 2016
First decision: February 4, 2017
Revised: February 11, 2017
Accepted: April 23, 2017
Article in press: April 24, 2017
Published online: May 18, 2017
Processing time: 181 Days and 15 Hours
Abstract
AIM

To retrospectively evaluate the diagnostic performance of free-breathing diffusion-weighted imaging (FB-DWI) with modified imaging parameter settings for detecting hepatocellular carcinomas (HCCs).

METHODS

Fifty-one patients at risk for HCC were scanned with both FB-DWI and respiratory-triggered DWI with the navigator echo respiratory-triggering technique (RT-DWI). Qualitatively, the sharpness of the liver contour, the image noise and the chemical shift artifacts on each DWI with b-values of 1000 s/mm2 were independently evaluated by three radiologists using 4-point scoring. We compared the image quality scores of each observer between the two DWI methods, using the Wilcoxon signed-rank test. Quantitatively, we compared the signal-to-noise ratios (SNRs) of the liver parenchyma and lesion-to-nonlesion contrast-to-noise ratios (CNRs) after measuring the signal intensity on each DWI with a b-factor of 1000 s/mm2. The average SNRs and CNRs between the two DWI methods were compared by the paired t-test. The detectability of HCC on each DWI was also analyzed by three radiologists. The detectability provided by the two DWI methods was compared using McNemar’s test.

RESULTS

For all observers, the averaged image quality scores of FB-DWI were: Sharpness of the liver contour [observer (Obs)-1, 3.08 ± 0.81; Obs-2, 2.98 ± 0.73; Obs-3, 3.54 ± 0.75], those of the distortion (Obs-1, 2.94 ± 0.50; Obs-2, 2.71 ± 0.70; Obs-3, 3.27 ± 0.53), and the chemical shift artifacts (Obs-1, 3.38 ± 0.60; Obs-2, 3.15 ± 1.07; Obs-3, 3.21 ± 0.85). The averaged image quality scores of RT-DWI were: Sharpness of the liver contour (Obs-1, 2.33 ± 0.65; Obs-2, 2.37 ± 0.74; Obs-3, 2.75 ± 0.81), distortion (Obs-1, 2.81 ± 0.56; Obs-2, 2.25 ± 0.74; Obs-3, 2.96 ± 0.71), and the chemical shift artifacts (Obs-1, 2.92 ± 0.59; Obs-2, 2.21 ± 0.85; Obs-3, 2.77 ± 1.08). All image quality scores of FB-DWI were significantly higher than those of RT-DWI (P < 0.05). The average SNR of the normal liver parenchyma by FB-DWI (11.0 ± 4.8) was not significantly different from that shown by RT-DWI (11.0 ± 5.0); nor were the lesion-to-nonlesion CNRs significantly different (FB-DWI, 21.4 ± 17.7; RT-DWI, 20.1 ± 15.1). For all three observers, the detectability of FB-DWI (Obs-1, 43.6%; Obs-2, 53.6%; and Obs-3, 45.0%) was significantly higher than that of RT-DWI (Obs-1, 29.1%; Obs-2, 43.6%; and Obs-3, 34.5%) (P < 0.05).

CONCLUSION

FB-DWI showed better image quality and higher detectability of HCC compared to RT-DWI, without significantly reducing the SNRs of the liver parenchyma and lesion-to-nonlesion CNRs.

Keywords: Diffusion weighted-imaging; Liver; Magnetic resonance imaging; Hepatocellular carcinoma; Free-breathing technique

Core tip: This retrospective study evaluated the image quality of free-breathing diffusion-weighted imaging (FB-DWI) of the liver and its diagnostic performance for hepatocellular carcinoma compared with respiratory-triggered DWI. The free-breathing technique is widely believed to be inappropriate for body DWI because motion artifact causes decreased image quality. However, after a modification of imaging parameters, FB-DWI showed better image quality without significantly reducing the signal-to-noise ratio of the normal liver parenchyma and the lesion-to-nonlesion contrast-to-noise ratio compared to respiratory-triggering-DWI. As a result, the improvement of the image quality of FB-DWI contributed to an increased rate of detection of hepatocellular carcinoma.