MR Fingerprinting for Liver Tissue Characterization: A Histopathologic Correlation Study

Shohei Fujita; Katsuhiro Sano; Gastao Cruz; Yuki Fukumura; Hideo Kawasaki; Issei Fukunaga; Yuichi Morita; Masami Yoneyama; Koji Kamagata; Osamu Abe; Kenichi Ikejima; René M. Botnar; Claudia Prieto; Shigeki Aoki

doi:10.1148/radiol.220736

MR Fingerprinting for Liver Tissue Characterization: A Histopathologic Correlation Study

Shohei Fujita, Katsuhiro Sano, Gastao Cruz, Yuki Fukumura, Hideo Kawasaki, Issei Fukunaga, Yuichi Morita, Masami Yoneyama, Koji Kamagata, Osamu Abe, Kenichi Ikejima, René M. Botnar, Claudia Prieto, Shigeki Aoki

Biomedical Engineering Department

Juntendo University School of Medicine

Research output: Contribution to journal › Article › peer-review

5 Citations (Scopus)

40 Downloads (Pure)

Abstract

Background Liver MR fingerprinting (MRF) enables simultaneous quantification of T1, T2, T2*, and proton density fat fraction (PDFF) maps in single breath-hold acquisitions. Histopathologic correlation studies are desired for its clinical use. Purpose To compare liver MRF-derived metrics with separate reference quantitative MRI in participants with diffuse liver disease, evaluate scan-rescan repeatability of liver MRF, and validate MRF-derived measurements for histologic grading of liver biopsies. Materials and Methods This prospective study included participants with diffuse liver disease undergoing MRI from July 2021 to January 2022. Participants underwent two-dimensional single-section liver MRF and separate reference quantitative MRI. Linear regression, Bland-Altman plots, and coefficients of variation were used to assess the bias and repeatability of liver MRF measurements. For participants undergoing liver biopsy, the association between mapping and histologic grading was evaluated by using the Spearman correlation coefficient. Results Fifty-six participants (mean age, 59 years ± 15 [SD]; 32 women) were included to compare mapping techniques and 23 participants were evaluated with liver biopsy (mean age, 52.7 years ± 12.7; 14 women). The linearity of MRF with reference measurements in participants with diffuse liver disease (R2 value) for T1, T2, T2*, and PDFF maps was 0.86, 0.88, 0.54, and 0.99, respectively. The overall coefficients of variation for repeatability in the liver were 3.2%, 5.5%, 7.1%, and 4.6% for T1, T2, T2*, and PDFF maps, respectively. MRF-derived metrics showed high diagnostic performance in differentiating moderate or severe changes from mild or no changes (area under the receiver operating characteristic curve for fibrosis, inflammation, steatosis, and siderosis: 0.62 [95% CI: 0.52, 0.62], 0.92 [95% CI: 0.88, 0.92], 0.97 [95% CI: 0.96, 0.97], and 0.74 [95% CI: 0.57, 0.74], respectively). Conclusion Liver MR fingerprinting provided repeatable T1, T2, T2*, and proton density fat fraction maps in high agreement with reference quantitative mapping and may correlate with pathologic grades in participants with diffuse liver disease.

Original language	English
Pages (from-to)	150-159
Number of pages	10
Journal	Radiology
Volume	306
Issue number	1
DOIs	https://doi.org/10.1148/radiol.220736
Publication status	Published - 1 Jan 2023

Access to Document

10.1148/radiol.220736

radiol.220736Final published version, 2.91 MB

Cite this

@article{49abab66748b4c3e99d63f453d691ce0,

title = "MR Fingerprinting for Liver Tissue Characterization: A Histopathologic Correlation Study",

abstract = "Background Liver MR fingerprinting (MRF) enables simultaneous quantification of T1, T2, T2*, and proton density fat fraction (PDFF) maps in single breath-hold acquisitions. Histopathologic correlation studies are desired for its clinical use. Purpose To compare liver MRF-derived metrics with separate reference quantitative MRI in participants with diffuse liver disease, evaluate scan-rescan repeatability of liver MRF, and validate MRF-derived measurements for histologic grading of liver biopsies. Materials and Methods This prospective study included participants with diffuse liver disease undergoing MRI from July 2021 to January 2022. Participants underwent two-dimensional single-section liver MRF and separate reference quantitative MRI. Linear regression, Bland-Altman plots, and coefficients of variation were used to assess the bias and repeatability of liver MRF measurements. For participants undergoing liver biopsy, the association between mapping and histologic grading was evaluated by using the Spearman correlation coefficient. Results Fifty-six participants (mean age, 59 years ± 15 [SD]; 32 women) were included to compare mapping techniques and 23 participants were evaluated with liver biopsy (mean age, 52.7 years ± 12.7; 14 women). The linearity of MRF with reference measurements in participants with diffuse liver disease (R2 value) for T1, T2, T2*, and PDFF maps was 0.86, 0.88, 0.54, and 0.99, respectively. The overall coefficients of variation for repeatability in the liver were 3.2%, 5.5%, 7.1%, and 4.6% for T1, T2, T2*, and PDFF maps, respectively. MRF-derived metrics showed high diagnostic performance in differentiating moderate or severe changes from mild or no changes (area under the receiver operating characteristic curve for fibrosis, inflammation, steatosis, and siderosis: 0.62 [95% CI: 0.52, 0.62], 0.92 [95% CI: 0.88, 0.92], 0.97 [95% CI: 0.96, 0.97], and 0.74 [95% CI: 0.57, 0.74], respectively). Conclusion Liver MR fingerprinting provided repeatable T1, T2, T2*, and proton density fat fraction maps in high agreement with reference quantitative mapping and may correlate with pathologic grades in participants with diffuse liver disease.",

author = "Shohei Fujita and Katsuhiro Sano and Gastao Cruz and Yuki Fukumura and Hideo Kawasaki and Issei Fukunaga and Yuichi Morita and Masami Yoneyama and Koji Kamagata and Osamu Abe and Kenichi Ikejima and Botnar, {Ren{\'e} M.} and Claudia Prieto and Shigeki Aoki",

note = "Funding Information: Supported by the Japan Radiological Society, Bayer Research Grant; Engineering and Physical Sciences Research Council (EP/V044087/1); Millennium Institute for Intelligent Healthcare Engineering iHEALTH (ICN2021_004) and Fondecyt (1210637 and 1210638). K.K. supported by a grant from JSPS Kakenhi (19K17244). Publisher Copyright: {\textcopyright} RSNA, 2022.",

year = "2023",

month = jan,

day = "1",

doi = "10.1148/radiol.220736",

language = "English",

volume = "306",

pages = "150--159",

journal = "Radiology",

issn = "1527-1315",

publisher = "Radiological Society of North America",

number = "1",

}

TY - JOUR

T1 - MR Fingerprinting for Liver Tissue Characterization

T2 - A Histopathologic Correlation Study

AU - Fujita, Shohei

AU - Sano, Katsuhiro

AU - Cruz, Gastao

AU - Fukumura, Yuki

AU - Kawasaki, Hideo

AU - Fukunaga, Issei

AU - Morita, Yuichi

AU - Yoneyama, Masami

AU - Kamagata, Koji

AU - Abe, Osamu

AU - Ikejima, Kenichi

AU - Botnar, René M.

AU - Prieto, Claudia

AU - Aoki, Shigeki

N1 - Funding Information: Supported by the Japan Radiological Society, Bayer Research Grant; Engineering and Physical Sciences Research Council (EP/V044087/1); Millennium Institute for Intelligent Healthcare Engineering iHEALTH (ICN2021_004) and Fondecyt (1210637 and 1210638). K.K. supported by a grant from JSPS Kakenhi (19K17244). Publisher Copyright: © RSNA, 2022.

PY - 2023/1/1

Y1 - 2023/1/1

N2 - Background Liver MR fingerprinting (MRF) enables simultaneous quantification of T1, T2, T2*, and proton density fat fraction (PDFF) maps in single breath-hold acquisitions. Histopathologic correlation studies are desired for its clinical use. Purpose To compare liver MRF-derived metrics with separate reference quantitative MRI in participants with diffuse liver disease, evaluate scan-rescan repeatability of liver MRF, and validate MRF-derived measurements for histologic grading of liver biopsies. Materials and Methods This prospective study included participants with diffuse liver disease undergoing MRI from July 2021 to January 2022. Participants underwent two-dimensional single-section liver MRF and separate reference quantitative MRI. Linear regression, Bland-Altman plots, and coefficients of variation were used to assess the bias and repeatability of liver MRF measurements. For participants undergoing liver biopsy, the association between mapping and histologic grading was evaluated by using the Spearman correlation coefficient. Results Fifty-six participants (mean age, 59 years ± 15 [SD]; 32 women) were included to compare mapping techniques and 23 participants were evaluated with liver biopsy (mean age, 52.7 years ± 12.7; 14 women). The linearity of MRF with reference measurements in participants with diffuse liver disease (R2 value) for T1, T2, T2*, and PDFF maps was 0.86, 0.88, 0.54, and 0.99, respectively. The overall coefficients of variation for repeatability in the liver were 3.2%, 5.5%, 7.1%, and 4.6% for T1, T2, T2*, and PDFF maps, respectively. MRF-derived metrics showed high diagnostic performance in differentiating moderate or severe changes from mild or no changes (area under the receiver operating characteristic curve for fibrosis, inflammation, steatosis, and siderosis: 0.62 [95% CI: 0.52, 0.62], 0.92 [95% CI: 0.88, 0.92], 0.97 [95% CI: 0.96, 0.97], and 0.74 [95% CI: 0.57, 0.74], respectively). Conclusion Liver MR fingerprinting provided repeatable T1, T2, T2*, and proton density fat fraction maps in high agreement with reference quantitative mapping and may correlate with pathologic grades in participants with diffuse liver disease.

AB - Background Liver MR fingerprinting (MRF) enables simultaneous quantification of T1, T2, T2*, and proton density fat fraction (PDFF) maps in single breath-hold acquisitions. Histopathologic correlation studies are desired for its clinical use. Purpose To compare liver MRF-derived metrics with separate reference quantitative MRI in participants with diffuse liver disease, evaluate scan-rescan repeatability of liver MRF, and validate MRF-derived measurements for histologic grading of liver biopsies. Materials and Methods This prospective study included participants with diffuse liver disease undergoing MRI from July 2021 to January 2022. Participants underwent two-dimensional single-section liver MRF and separate reference quantitative MRI. Linear regression, Bland-Altman plots, and coefficients of variation were used to assess the bias and repeatability of liver MRF measurements. For participants undergoing liver biopsy, the association between mapping and histologic grading was evaluated by using the Spearman correlation coefficient. Results Fifty-six participants (mean age, 59 years ± 15 [SD]; 32 women) were included to compare mapping techniques and 23 participants were evaluated with liver biopsy (mean age, 52.7 years ± 12.7; 14 women). The linearity of MRF with reference measurements in participants with diffuse liver disease (R2 value) for T1, T2, T2*, and PDFF maps was 0.86, 0.88, 0.54, and 0.99, respectively. The overall coefficients of variation for repeatability in the liver were 3.2%, 5.5%, 7.1%, and 4.6% for T1, T2, T2*, and PDFF maps, respectively. MRF-derived metrics showed high diagnostic performance in differentiating moderate or severe changes from mild or no changes (area under the receiver operating characteristic curve for fibrosis, inflammation, steatosis, and siderosis: 0.62 [95% CI: 0.52, 0.62], 0.92 [95% CI: 0.88, 0.92], 0.97 [95% CI: 0.96, 0.97], and 0.74 [95% CI: 0.57, 0.74], respectively). Conclusion Liver MR fingerprinting provided repeatable T1, T2, T2*, and proton density fat fraction maps in high agreement with reference quantitative mapping and may correlate with pathologic grades in participants with diffuse liver disease.

UR - http://www.scopus.com/inward/record.url?scp=85144589631&partnerID=8YFLogxK

U2 - 10.1148/radiol.220736

DO - 10.1148/radiol.220736

M3 - Article

C2 - 36040337

AN - SCOPUS:85144589631

SN - 1527-1315

VL - 306

SP - 150

EP - 159

JO - Radiology

JF - Radiology

IS - 1

ER -

MR Fingerprinting for Liver Tissue Characterization: A Histopathologic Correlation Study

Abstract

Access to Document

Other files and links

Fingerprint

Cite this