TY - JOUR
T1 - Synthetic magnetic resonance imaging for primary prostate cancer evaluation
T2 - Diagnostic potential of a non-contrast-enhanced bi-parametric approach enhanced with relaxometry measurements
AU - Arita, Yuki
AU - Akita, Hirotaka
AU - Fujiwara, Hirokazu
AU - Hashimoto, Masahiro
AU - Shigeta, Keisuke
AU - Kwee, Thomas C.
AU - Yoshida, Soichiro
AU - Kosaka, Takeo
AU - Okuda, Shigeo
AU - Oya, Mototsugu
AU - Jinzaki, Masahiro
N1 - Funding Information:
The authors thank Mr. Koshi Okabe and Ms. Sari Motomatsu for their help with data collection.
Publisher Copyright:
© 2022
PY - 2022/1
Y1 - 2022/1
N2 - Purpose: Bi-parametric magnetic resonance imaging (bpMRI) with diffusion-weighted images has wide utility in diagnosing clinically significant prostate cancer (csPCa). However, bpMRI yields more false-negatives for PI-RADS category 3 lesions than multiparametric (mp)MRI with dynamic-contrast-enhanced (DCE)-MRI. We investigated the utility of synthetic MRI with relaxometry maps for bpMRI-based diagnosis of csPCa. Methods: One hundred and five treatment-naïve patients who underwent mpMRI and synthetic MRI before prostate biopsy for suspected PCa between August 2019 and December 2020 were prospectively included. Three experts and three basic prostate radiologists evaluated the diagnostic performance of conventional bpMRI and synthetic bpMRI for csPCa. PI-RADS version 2.1 category 3 lesions were identified by consensus, and relaxometry measurements (T1-value, T2-value, and proton density [PD]) were performed. The diagnostic performance of relaxometry measurements for PI-RADS category 3 lesions in peripheral zone was compared with that of DCE-MRI. Histopathological evaluation results were used as the reference standard. Statistical analysis was performed using the areas under the receiver operating characteristic curve (AUC) and McNemar test. Results: In 102 patients without significant MRI artefacts, the diagnostic performance of conventional bpMRI was not significantly different from that of synthetic bpMRI for all readers (p = 0.11–0.79). The AUCs of the combination of T1-value, T2-value, and PD (T1 + T2 + PD) for csPCa in peripheral zone for PI-RADS category 3 lesions were 0.85 for expert and 0.86 for basic radiologists, with no significant difference between T1 + T2 + PD and DCE-MRI for both expert and basic radiologists (p = 0.29–0.45). Conclusion: Synthetic MRI with relaxometry maps shows promise for contrast media-free evaluation of csPCa.
AB - Purpose: Bi-parametric magnetic resonance imaging (bpMRI) with diffusion-weighted images has wide utility in diagnosing clinically significant prostate cancer (csPCa). However, bpMRI yields more false-negatives for PI-RADS category 3 lesions than multiparametric (mp)MRI with dynamic-contrast-enhanced (DCE)-MRI. We investigated the utility of synthetic MRI with relaxometry maps for bpMRI-based diagnosis of csPCa. Methods: One hundred and five treatment-naïve patients who underwent mpMRI and synthetic MRI before prostate biopsy for suspected PCa between August 2019 and December 2020 were prospectively included. Three experts and three basic prostate radiologists evaluated the diagnostic performance of conventional bpMRI and synthetic bpMRI for csPCa. PI-RADS version 2.1 category 3 lesions were identified by consensus, and relaxometry measurements (T1-value, T2-value, and proton density [PD]) were performed. The diagnostic performance of relaxometry measurements for PI-RADS category 3 lesions in peripheral zone was compared with that of DCE-MRI. Histopathological evaluation results were used as the reference standard. Statistical analysis was performed using the areas under the receiver operating characteristic curve (AUC) and McNemar test. Results: In 102 patients without significant MRI artefacts, the diagnostic performance of conventional bpMRI was not significantly different from that of synthetic bpMRI for all readers (p = 0.11–0.79). The AUCs of the combination of T1-value, T2-value, and PD (T1 + T2 + PD) for csPCa in peripheral zone for PI-RADS category 3 lesions were 0.85 for expert and 0.86 for basic radiologists, with no significant difference between T1 + T2 + PD and DCE-MRI for both expert and basic radiologists (p = 0.29–0.45). Conclusion: Synthetic MRI with relaxometry maps shows promise for contrast media-free evaluation of csPCa.
KW - Biomarkers
KW - Contrast agents
KW - Magnetic resonance imaging
KW - Multiparametric Magnetic Resonance Imaging
KW - Primary prostate cancer
UR - http://www.scopus.com/inward/record.url?scp=85124522146&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85124522146&partnerID=8YFLogxK
U2 - 10.1016/j.ejro.2022.100403
DO - 10.1016/j.ejro.2022.100403
M3 - Article
AN - SCOPUS:85124522146
SN - 2352-0477
VL - 9
JO - European Journal of Radiology Open
JF - European Journal of Radiology Open
M1 - 100403
ER -
