Share this content in WeChat
Special Focus
Novel deep learning-based T2-weighted imaging of the prostate provides superior image quality
KE Zan  LI Liang  SONG Xinyang  WEN Zhi  GAO Yufan  LIU Weiyin  QUAN Guangnan  ZHA Yunfei 


[Abstract] Objective To introduce a novel deep learning-based reconstruction (DLR, which is now commercially available as AIRTM Recon DL, GE Healthcare) T2-weighted imaging (T2WIDL) sequence in prostate MRI and investigate its image quality and diagnostic confidence compared to conventional T2-weighted imaging (T2WIC).Materials and Methods Seventy-eight patients who underwent prostate MRI examinations (T2WIC and T2WIDL with the same parameters) were included in this retrospective study. For the qualitative and diagnostic confidence evaluation, double-blinded evaluation was performed by both three- and seven-year experienced radiologists according to the Likert Scale (5=excellent, 1=very poor), and then the difference among the scores were evaluated using Wilcoxon test and the intra- /inter- observer agreement were evaluated using κ statistics. The evaluation indicators of T2WI image quality and diagnostic confidence including: prostate capsule, lesion contrast and edge sharpness, anatomical details (urethra, zone of prostate, seminal vesicle), skeleton and muscle clarity, overall image quality, lesion location and morphology, lesion is benign or malignant. In addition, the time spent by two radiologists browsing each set of images was recorded respectively, and the paired t test was used for statistical analysis. As for quantitative evaluation, the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) measured between each prostate lesion on the MR images acquired with different sequences were analyzed, paired t test and Mann-Whitney U test were used for statistical analysis.Results Seventy-eight patients at the mean age of (67.1±9.9) years were included in this retrospective study. Based on the subjective scoring criteria, overall image quality scores were rated significantly superior by both readers with (4.6±0.6) and (4.3±0.7) on T2WIDL compared to (3.4±0.7) and (3.0±0.8) on T2WIC (P<0.05). For T2WIDL, the score consistency ranged from 0.6 to 0.8; there were significant differences in the scores between the two readers only for anatomical details and overall image quality (P<0.05). Besides, overall diagnostic confidence scores also were rated significantly superior by both readers with (4.8±0.3) and (4.8±0.4) on T2WIDL compared to (3.8±0.4) and (3.7±0.5) on T2WIC (P<0.05), with fewer time to spend. Based on objective evaluation, SNR and CNR of T2WIDL were higher than those of T2WIC, and the differences were statistically significant (P<0.05). The SNR of T2WIC and T2WIDL in benign and malignant lesions were (12.4±2.4), (28.7±8.1) and (10.1±1.8), (27.7±5.4), respectively, with significant differences (P<0.01). There was no significant difference in CNR between benign and malignant lesions with and without DL (P>0.05).Conclusions The prostate T2WIDL images have high subjective rating scores, clearer lesion contrast, high SNR and CNR. In addition, the radiologists had more diagnostic confidence in T2WIDL image with less diagnostic time. Therefore, the novel DLR technique is helpful to improve the image quality of prostate T2WI within the same scanning time, which provides a more accurate imaging basis for clinical diagnosis and treatment.
[Keywords] prostate;prostate cancer;deep learning;magnetic resonance imaging;image quality;signal-to-noise ratio;contrast-to-noise ratio

KE Zan1   LI Liang1   SONG Xinyang2   WEN Zhi1   GAO Yufan1   LIU Weiyin3   QUAN Guangnan3   ZHA Yunfei1*  

1 Department of Radiology, Renmin Hospital of Wuhan University, Wuhan 430060, China

2 Department of Radiology, Xiangyang No.1 People's Hospital, Hubei University of Medicine, Xiangyang 441000, China

3 GE Healthcare (China) Co., Ltd., Beijing 100176, China

Corresponding author: Zha YF, E-mail:

Conflicts of interest   None.

Received  2022-09-14
Accepted  2022-11-29
DOI: 10.12015/issn.1674-8034.2023.05.009

PECORARO M, TURKBEY B, PURYSKO A S, et al. Diagnostic accuracy and observer agreement of the MRI prostate imaging for recurrence reporting assessment score[J]. Radiology, 2022, 304(2): 342-350. DOI: 10.1148/radiol.212252">10.1148/radiol.212252">10.1148/radiol.212252.
PANEBIANCO V, VILLEIRS G, WEINREB J C, et al. Prostate magnetic resonance imaging for local recurrence reporting (PI-RR): international consensus-based guidelines on multiparametric magnetic resonance imaging for prostate cancer recurrence after radiation therapy and radical prostatectomy[J]. Eur Urol Oncol, 2021, 4(6): 868-876. DOI: 10.1016/j.euo.2021.01.003">10.1016/j.euo.2021.01.003">10.1016/j.euo.2021.01.003.
ZHAO J, KADER A, MANGAROVA D B, et al. Dynamic Contrast-Enhanced MRI of Prostate Lesions of Simultaneous [68Ga] Ga-PSMA-11 PET/MRI: Comparison between Intraprostatic Lesions and Correlation between Perfusion Parameters[J/OL]. Cancers (Basel), 2021, 13(6): 1404 [2022-10-18]. DOI: 10.3390/cancers13061404">10.3390/cancers13061404">10.3390/cancers13061404.
WEINREB J C, BARENTSZ J O, CHOYKE P L, et al. PI-RADS prostate imaging - reporting and data system: 2015, version 2[J]. Eur Urol, 2016, 69(1): 16-40. DOI: 10.1016/j.eururo.2015.08.052">10.1016/j.eururo.2015.08.052">10.1016/j.eururo.2015.08.052.
TURKBEY B, ROSENKRANTZ A B, HAIDER M A, et al. Prostate Imaging Reporting and Data System Version 2.1: 2019 Update of Prostate Imaging Reporting and Data System Version 2[J]. Eur Urol, 2019, 76(3): 340-351. DOI: 10.1016/j.eururo.2019.02.033">10.1016/j.eururo.2019.02.033">10.1016/j.eururo.2019.02.033.
TAKAGI H, KADOYA N, KAJIKAWA T, et al. Multi-atlas-based auto-segmentation for prostatic urethra using novel prediction of deformable image registration accuracy[J]. Med Phys, 2020, 47(7): 3023-3031. DOI: 10.1002/mp.14154">10.1002/mp.14154">10.1002/mp.14154.
CUOCOLO R, CIPULLO M B, STANZIONE A, et al. Machine learning applications in prostate cancer magnetic resonance imaging[J/OL]. Eur Radiol Exp, 2019, 3(1): 35 [2022-10-18]. DOI: 10.1186/s41747-019-0109-2">10.1186/s41747-019-0109-2">10.1186/s41747-019-0109-2.
HAMZEH O, ALKHATEEB A, ZHENG J, et al. Prediction of tumor location in prostate cancer tissue using a machine learning system on gene expression data[J/OL]. BMC Bioinformatics, 2020, 21(Suppl 2): 78 [2022-10-18]. DOI: 10.1186/s12859-020-3345-9">10.1186/s12859-020-3345-9">10.1186/s12859-020-3345-9.
TURKBEY B. Better image quality for diffusion-weighted MRI of the prostate using deep learning[J]. Radiology, 2022, 303(2): 382-383. DOI: 10.1148/radiol.212078">10.1148/radiol.212078">10.1148/radiol.212078.
LUNDERVOLD A S, LUNDERVOLD A. An overview of deep learning in medical imaging focusing on MRI[J]. Z Med Phys, 2019, 29(2): 102-127. DOI: 10.1016/j.zemedi.2018.11.002">10.1016/j.zemedi.2018.11.002">10.1016/j.zemedi.2018.11.002.
KIDOH M, SHINODA K, KITAJIMA M, et al. Deep learning based noise reduction for brain MR imaging: tests on phantoms and healthy volunteers[J]. Magn Reson Med Sci, 2020, 19(3): 195-206. DOI: 10.2463/">10.2463/">10.2463/
GASSENMAIER S, AFAT S, NICKEL M D, et al. Accelerated T2-Weighted TSE Imaging of the Prostate Using Deep Learning Image Reconstruction: A Prospective Comparison with Standard T2-Weighted TSE Imaging[J/OL]. Cancers (Basel), 2021, 13(14): 3593 [2022-10-18]. DOI: 10.3390/cancers13143593">10.3390/cancers13143593">10.3390/cancers13143593.
KIM M, KIM H S, KIM H J, et al. Thin-slice pituitary MRI with deep learning-based reconstruction: diagnostic performance in a postoperative setting[J]. Radiology, 2021, 298(1): 114-122. DOI: 10.1148/radiol.2020200723">10.1148/radiol.2020200723">10.1148/radiol.2020200723.
PETERS R D, HARRIS H, LAWSON S. The clinical benefits of AIRTM Recon DL for MR image reconstruction[EB/OL]. [2022-11-07].
LEBEL R M. Performance characterization of a novel deeplearning-based MR image reconstruction pipeline[EB/OL]. [2022-11-07].
SUN S, TAN E T, MINTZ D N, et al. Evaluation of deep learning reconstructed high-resolution 3D lumbar spine MRI[J]. Eur Radiol, 2022, 32(9): 6167-6177. DOI: 10.1007/s00330-022-08708-4">10.1007/s00330-022-08708-4">10.1007/s00330-022-08708-4.
KANIEWSKA M, DEININGER-CZERMAK E, GETZMANN J M, et al. Application of deep learning-based image reconstruction in MR imaging of the shoulder joint to improve image quality and reduce scan time[J/OL]. Eur Radiol, 2022 [2022-11-13]. DOI: 10.1007/s00330-022-09151-1">10.1007/s00330-022-09151-1">10.1007/s00330-022-09151-1.
WANG X Z, MA J F, BHOSALE P, et al. Novel deep learning-based noise reduction technique for prostate magnetic resonance imaging[J]. Abdom Radiol (NY), 2021, 46(7): 3378-3386. DOI: 10.1007/s00261-021-02964-6">10.1007/s00261-021-02964-6">10.1007/s00261-021-02964-6.
KIM E H, CHOI M H, LEE Y J, et al. Deep learning-accelerated T2-weighted imaging of the prostate: Impact of further acceleration with lower spatial resolution on image quality[J/OL]. Eur J Radiol, 2021, 145: 110012 [2022-10-18]. https://pubmed.ncbi.nlm.nih.-gov/34753082. DOI: 10.1016/j.ejrad.2021.110012">10.1016/j.ejrad.2021.110012">10.1016/j.ejrad.2021.110012.
YANG Q, ZOU L Y, LIU Z, et al. Application of a dedicated surface coil in thyroid MRI provides superior image quality[J]. Chin J Magn Reson Imaging, 2021, 12(2): 57-61. DOI: 10.12015/issn.1674-8034.2021.02.013">10.12015/issn.1674-8034.2021.02.013">10.12015/issn.1674-8034.2021.02.013.
GIGANTI F, ALLEN C, EMBERTON M, et al. Prostate imaging quality (PI-QUAL): a new quality control scoring system for multiparametric magnetic resonance imaging of the prostate from the PRECISION trial[J]. Eur Urol Oncol, 2020, 3(5): 615-619. DOI: 10.1016/j.euo.2020.06.007">10.1016/j.euo.2020.06.007">10.1016/j.euo.2020.06.007.
UEDA T, OHNO Y, YAMAMOTO K, et al. Deep learning reconstruction of diffusion-weighted MRI improves image quality for prostatic imaging[J]. Radiology, 2022, 303(2): 373-381. DOI: 10.1148/radiol.204097">10.1148/radiol.204097">10.1148/radiol.204097.
LI L, WANG L, DENG M, et al. Feasibility study of 3-T DWI of the prostate: readout-segmented versus single-shot echo-planar imaging[J]. AJR Am J Roentgenol, 2015, 205(1): 70-76. DOI: 10.2214/AJR.14.13489">10.2214/AJR.14.13489">10.2214/AJR.14.13489.
TURKBEY B, CHOYKE P L. Multiparametric MRI and prostate cancer diagnosis and risk stratification[J]. Curr Opin Urol, 2012, 22(4): 310-315. DOI: 10.1097/MOU.0b013e32835481c2">10.1097/MOU.0b013e32835481c2">10.1097/MOU.0b013e32835481c2.
CHOI M H, LEE Y J, JUNG S E. Tracking changes in clinical practice patterns following prebiopsy biparametric prostate MRI[J]. Acad Radiol, 2020, 27(9): 1255-1260. DOI: 10.1016/j.acra.2019.10.033">10.1016/j.acra.2019.10.033">10.1016/j.acra.2019.10.033.
WANG L, LI Q B, VARGAS H A. China interpretation of prostate imaging-reporting and data system (PI-RADS V2.1) guideline for prostate cancer management[J]. Chin J Radiol, 2020, 54(4): 273-278. DOI: 10.3760/cma.j.cn112149-20190429-00382">10.3760/cma.j.cn112149-20190429-00382">10.3760/cma.j.cn112149-20190429-00382.
PADHANI A R, BARENTSZ J, VILLEIRS G, et al. PI-RADS steering committee: the PI-RADS multiparametric MRI and MRI-directed biopsy pathway[J]. Radiology, 2019, 292(2): 464-474. DOI: 10.1148/radiol.2019182946">‍10.1148/radiol.2019182946">10.1148/radiol.2019182946.
SCHELB P, KOHL S, RADTKE J P, et al. Classification of cancer at prostate MRI: deep learning versus clinical PI-RADS assessment[J]. Radiology, 2019, 293(3): 607-617. DOI: 10.1148/radiol.2019190938">10.1148/radiol.2019190938">10.1148/radiol.2019190938.
ZHONG X R, CAO R M, SHAKERI S, et al. Deep transfer learning-based prostate cancer classification using 3 Tesla multi-parametric MRI[J]. Abdom Radiol (NY), 2019, 44(6): 2030-2039. DOI: 10.1007/s00261-018-1824-5">‍10.1007/s00261-018-1824-5">10.1007/s00261-018-1824-5.
SHENG R F, ZHENG L Y, JIN K P, et al. Single-breath-hold T2WI liver MRI with deep learning-based reconstruction: A clinical feasibility study in comparison to conventional multi-breath-hold T2WI liver MRI[J/OL]. Magn Reson Imaging, 2021, 81: 75-81 [2022-11-06]. DOI: 10.1016/j.mri.2021.06.014">10.1016/j.mri.2021.06.014">10.1016/j.mri.2021.06.014.
CHAUDHARI A S, FANG Z N, KOGAN F, et al. Super-resolution musculoskeletal MRI using deep learning[J]. Magn Reson Med, 2018, 80(5): 2139-2154. DOI: 10.1002/mrm.27178">10.1002/mrm.27178">10.1002/mrm.27178.
GASSENMAIER S, AFAT S, NICKEL D, et al. Deep learning-accelerated T2-weighted imaging of the prostate: Reduction of acquisition time and improvement of image quality[J/OL]. Eur J Radiol, 2021, 137: 109600 [2022-11-06]. DOI: 10.1016/j.ejrad.2021.109600.

PREV Feasibility study of deep learning reconstruction in the clinical application of MRI in bladder cancer
NEXT Value of deep learning reconstruction in optimizing prostate MR T2-weighted imaging scanning time and imaging quality

Tel & Fax: +8610-67113815    E-mail: