Share：
Share this content in WeChat
X
[Chinese] [PDF] 11 1
• Clinical Article •
Super-resolution reconstruction technique enhances the diagnostic efficacy of deep learning-based prediction of lymphvascular space invasion in endometrial cancer
YI Qinqin  WANG Ying  ZOU Sisi  GONG Jingshan 

Cite this article as: YI Q Q, WANG Y, ZOU S S, et al. Super-resolution reconstruction technique enhances the diagnostic efficacy of deep learning-based prediction of lymphvascular space invasion in endometrial cancer[J]. Chin J Magn Reson Imaging, 2025, 16(10): 83-88. DOI:10.12015/issn.1674-8034.2025.10.013.


[Abstract] Objective To evaluate whether super-resolution reconstruction technology can improve the diagnostic efficacy of deep learning in predicting lymphovascular space invasion (LVSI) in endometrial cancer.Materials and Methods This retrospective study enrolled 406 patients randomly split into training (n = 325) and validation (n = 81) sets (8∶2 ratio). Super-resolution reconstruction was performed on conventional T2-weighted imaging (T2WI) to obtain super high-resolution T2WI (SRT2). Deep learning models were developed based on both conventional T2WI and SRT2 images to predict LVSI status in endometrial cancer. The models were subsequently validated in validation set, and their diagnostic performance was compared across the training and validation sets. Using pathological diagnosis as the gold standard, the evaluation metrics included the area under the curve (AUC), sensitivity, and specificity, with model differences compared using DeLong's test.Results In both the training and validation sets, the deep learning model based on conventional T2WI demonstrated AUC values (95% confidence interval) of 0.792 (0.733 to 0.851) and 0.759 (0.649 to 0.870), with sensitivities of 77.50% and 68.18%, and specificities of 77.08% and 80.67%, respectively. The model utilizing SRT2 achieved AUCs of 0.897 (0.852 to 0.943) and 0.899 (0.819 to 0.980), sensitivities of 87.80% and 86.40%, and specificities of 88.45% and 89.20%. Statistically significant differences between the two models were observed in both sets (P < 0.05), indicating superior performance of the SRT2-based deep learning model.Conclusions Super-resolution reconstruction technology has the potential to enhance the diagnostic efficacyof preoperative prediction of LVSI in endometrial cancer by improving image quality.
[Keywords] endometrial cancer；deep learning；super-resolution reconstruction；magnetic resonance imaging；lymphovascular space invasion；radiomics；preoperative prediction

YI Qinqin1   WANG Ying1   ZOU Sisi2   GONG Jingshan1*  

1 Department of Radiology, Shenzhen People's Hospital (The Second Clinical Medical College of Jinan University, The First Affiliated Hospital of Southern University of Science and Technology), Shenzhen 518000, China

2 Department of Radiology, Shenzhen Second People's Hospital, Shenzhen 518000, China

Corresponding author: GONG J S, E-mail: jshgong@sina.com

Conflicts of interest   None.

Received  2025-04-22
Accepted  2025-09-14
DOI: 10.12015/issn.1674-8034.2025.10.013
Cite this article as: YI Q Q, WANG Y, ZOU S S, et al. Super-resolution reconstruction technique enhances the diagnostic efficacy of deep learning-based prediction of lymphvascular space invasion in endometrial cancer[J]. Chin J Magn Reson Imaging, 2025, 16(10): 83-88. DOI:10.12015/issn.1674-8034.2025.10.013.

[1]
HAN Y, SUN M Y. Advances in MRI studies of lymphvascular space invasion in endometrial carcinoma[J]. Chin J Magn Reson Imag, 2024, 15(7): 216-220, 234. DOI: 10.12015/issn.1674-8034.2024.07.036.
[2]
CHEN X J, ZHANG J F, LU W, et al. Chinese expert consensus on quality control and quality indicators forsurgical treatment of endometrial carcinoma (2024 edition)[J]. Chin J Pract Gynecol Obstet, 2024, 40(6): 626-637. DOI: 10.19538/j.fk2024060113.
[3]
TORTORELLA L, RESTAINO S, ZANNONI G F, et al. Substantial lymph-vascular space invasion (LVSI) as predictor of distant relapse and poor prognosis in low-risk early-stage endometrial cancer[J/OL]. J Gynecol Oncol, 2021, 32(2): e11 [2025-02-10]. https://pubmed.ncbi.nlm.nih.gov/33470061. DOI: 10.3802/jgo.2021.32.e11.
[4]
LIN Y X, ZHANG Y H, ZHANG G M, et al. Prognostic value of lymphovascular space invasion evaluated by immunohistochemistry and stratification with a semi-quantitative scoring system for early-stage endometrial cancer[J]. J Army Med Univ, 2022, 44(13): 1370-1377. DOI: 10.16016/j.2097-0927.202112273.
[5]
LEO E, STANZIONE A, MIELE M, et al. Artificial intelligence and radiomics for endometrial cancer MRI: exploring the whats, whys and hows[J/OL]. J Clin Med, 2023, 13(1): 226 [2025-02-10]. https://pubmed.ncbi.nlm.nih.gov/38202233. DOI: 10.3390/jcm13010226.
[6]
LONG L, LIU M L, DENG X J, et al. Tumor stiffness measurement at multifrequency MR elastography to predict lymphovascular space invasion in endometrial cancer[J/OL]. Radiology, 2024, 311(3): e232242 [2025-02-10]. https://pubmed.ncbi.nlm.nih.gov/38832881. DOI: 10.1148/radiol.232242.
[7]
LIU D, HUANG J Y, ZHANG Y F, et al. Multimodal MRI-based radiomics models for the preoperative prediction of lymphovascular space invasion of endometrial carcinoma[J/OL]. BMC Med Imaging, 2024, 24(1): 252 [2025-02-13]. https://pubmed.ncbi.nlm.nih.gov/39304802. DOI: 10.1186/s12880-024-01430-1.
[8]
CHEN J Y, WANG X R, LV H Y, et al. Development and external validation of a clinical-radiomics nomogram for preoperative prediction of LVSI status in patients with endometrial carcinoma[J]. J Cancer Res Clin Oncol, 2023, 149(15): 13943-13953. DOI: 10.1007/s00432-023-05044-y.
[9]
TSHALIBE P, ADRIGWE J, LUCAS S. Clinicians' perspective of picture archiving and communication systems at Charlotte Maxeke Johannesburg Academic Hospital[J/OL]. SA J Radiol, 2023, 27(1): 2578 [2025-02-13]. https://pubmed.ncbi.nlm.nih.gov/37292419. DOI: 10.4102/sajr.v27i1.2578.
[10]
ZHOU M, ZHANG M M, XIAO W B. Preliminary research on quality control of PACS image display device[J]. J Pract Radiol, 2005(9): 977-979, 990. DOI: 10.3969/j.issn.1002-1671.2005.09.024.
[11]
ZHANG M, XIA C C, TANG J, et al. Evaluation of high-resolution pituitary dynamic contrast-enhanced MRI using deep learning-based compressed sensing and super-resolution reconstruction[J]. Eur Radiol, 2025, 35(10): 5922-5934. DOI: 10.1007/s00330-025-11574-5.
[12]
ZHENG Q F, JIN Y, JIANG Z W, et al. Medical image super-resolution reconstruction based on back projection regression network[J]. J Comput Aided Des Comput Graph, 2024, 36(12): 1968-1977. DOI: 10.3724/SP.J.1089.2024.20113.
[13]
LI Z Y, LI Z H, LI H X, et al. Enhance the Image: Super Resolution using Artificial Intelligence in MRI[M/OL]. Arxiv, 2024 [2025-02-15]. https://arxiv.org/pdf/2406.13625.
[14]
OSCANOA J A, MIDDIONE M J, ALKAN C, et al. Deep learning-based reconstruction for cardiac MRI: a review[J/OL]. Bioengineering (Basel), 2023, 10(3): 334 [2025-02-15]. https://pubmed.ncbi.nlm.nih.gov/36978725. DOI: 10.3390/bioengineering10030334.
[15]
BIAN Y Y, WANG L, LI J, et al. Quantitative ischemic lesions of portable low-field strength MRI using deep learning-based super-resolution[J]. Stroke, 2025, 56(7): 1843-1852. DOI: 10.1161/STROKEAHA.124.050540.
[16]
ÇELIK B, MIKAEILI M, GENÇ M Z, et al. Can super resolution via deep learning improve classification accuracy in dental radiography [J]. Dentomaxillofac Radiol, 2025, 54(6): 473-487. DOI: 10.1093/dmfr/twaf029.
[17]
TAKAFUJI M, KITAGAWA K, MIZUTANI S, et al. Super-resolution deep learning reconstruction for improved image quality of coronary CT angiography[J/OL]. Radiol Cardiothorac Imaging, 2023, 5(4): e230085 [2025-02-25]. https://pubmed.ncbi.nlm.nih.gov/37693207. DOI: 10.1148/ryct.230085.
[18]
CHENG Z B, XIE L L, FENG C X, et al. Super-resolution acquisition and reconstruction for cone-beam SPECT with low-resolution detector[J/OL]. Comput Methods Programs Biomed, 2022, 217: 106683 [2025-02-25]. https://pubmed.ncbi.nlm.nih.gov/35150999. DOI: 10.1016/j.cmpb.2022.106683.
[19]
OTA J, UMEHARA K, KERSHAW J, et al. Super-resolution generative adversarial networks with static T2*WI-based subject-specific learning to improve spatial difference sensitivity in fMRI activation[J/OL]. Sci Rep, 2022, 12: 10319 [2025-02-25]. https://pubmed.ncbi.nlm.nih.gov/35725788. DOI: 10.1038/s41598-022-14421-5.
[20]
CHEN S, MA D, LEE S, et al. Segmentation-guided domain adaptation and data harmonization of multi-device retinal optical coherence tomography using cycle-consistent generative adversarial networks[J/OL]. Comput Biol Med, 2023, 159: 106595 [2025-03-10]. https://pubmed.ncbi.nlm.nih.gov/37087780. DOI: 10.1016/j.compbiomed.2023.106595.
[21]
HSU W W, GUO J M, CHEN C Y, et al. Fall detection with the spatial-temporal correlation encoded by a sequence-to-sequence denoised GAN[J/OL]. Sensors (Basel), 2022, 22(11): 4194 [2025-03-10]. https://pubmed.ncbi.nlm.nih.gov/35684812. DOI: 10.3390/s22114194.
[22]
MAHMOUD M, KANG H S. GANMasker: A two-stage generative adversarial network for high-quality face mask removal[J/OL]. Sensors, 2023, 23(16): 7094 [2025-03-15]. https://pubmed.ncbi.nlm.nih.gov/37631631. DOI: 10.3390/s23167094.
[23]
JIANG Z. Research on generative adversarial networks based on improved loss functions[D]. Wuhan: Wuhan University of Science and Technology, 2023. DOI: 10.27380/d.cnki.gwkju.2023.000374.
[24]
ELMEZAIN M, MAHMOUD A, MOSA D T, et al. Brain tumor segmentation using deep capsule network and latent-dynamic conditional random fields[J/OL]. J Imaging, 2022, 8(7): 190 [2025-03-15]. https://pubmed.ncbi.nlm.nih.gov/35877634. DOI: 10.3390/jimaging8070190.
[25]
DOVROU A, NIKIFORAKI K, ZARIDIS D, et al. A segmentation-based method improving the performance of N4 bias field correction on T2weighted MR imaging data of the prostate[J]. Magn Reson Imaging, 2023, 101: 1-12. DOI: 10.1016/j.mri.2023.03.012.
[26]
GUO Y J, YIN R, HAN J Q, et al. Research on the deep learning model based on the combination of intratumoral and peritumoral dynamic contrast-enhanced MRI for predicting axillary lymph node metastasis in breast cancer[J]. Journal of Practical Radiology, 2024.40(6): 907-912. DOI: 10.3969/j.issn.1002-1671.2024.06.011.
[27]
SUN J H, YANG Y, XUN G X, et al. Scheduling hyperparameters to improve generalization: from centralized SGD to asynchronous SGD[J/OL]. ACM Trans Knowl Discov Data, 2023.17(2): 1-37. DOI: 10.1145/3544782.
[28]
HEO T, LEE N K, KIM S, et al. Deep learning reconstruction of diffusion-weighted imaging with single-shot echo-planar imaging in endometrial cancer: a comparison with multi-shot echo-planar imaging[J/OL]. Abdom Radiol, 2025: [2025-05-07]. https://pubmed.ncbi.nlm.nih.gov/40249551. DOI: 10.1007/s00261-025-04955-3.
[29]
WANG Z C, HU Y, CAI J, et al. Potential value of novel multiparametric MRI radiomics for preoperative prediction of microsatellite instability and Ki-67 expression in endometrial cancer[J/OL]. Sci Rep, 2025, 15: 3226 [2025-05-07]. https://pubmed.ncbi.nlm.nih.gov/39863695. DOI: 10.1038/s41598-025-87966-w.
[30]
PENG Y J, LIU X W, TANG X, et al. Preoperative predicting lymphov-ascular space invasion in endometrial carcinoma by nomogram based on mpMRI radiomics[J]. Chin J Magn Reson Imag, 2022, 13(7): 61-67. DOI: 10.12015/issn.1674-8034.2022.07.011.
[31]
FENG Y, LAN X L. Introduction to imaging genomics[J]. Chin J Nucl Med Mol Imag, 2023, 43(1): 55-60. DOI: 10.3760/cma.j.cn321828-20211130-00427.FENG Y, LAN X L. Introduction to radiomics[J]. Chinese Journal of Nuclear Medicine and Molecular Imaging, 2023.43(1): 55-60. DOI: 10.3760/cma.j.cn321828-20211130-00427.
[32]
HE Y M, LUO J W, ZHANG T, et al. The value of radiomics model based on MR hyperresolution reconstructed images in preoperative prediction of lymphatic vascular infiltration in early cervical cancer[J]. Radiol Pract, 2024, 39(9): 1213-1220. DOI: 10.13609/j.cnki.1000-0313.2024.09.015.
[33]
ZHANG J D, JIA Y Y, ZHU H Z, et al. 3D-MRI super-resolution algorithm fusing attention and dilated encoder-decoder[J]. Comput Eng Appl, 2024, 60(13): 228-236. DOI: 10.3778/j.issn.1002-8331.2304-0099.
[34]
XIA H, LÜ H F, LUO J, et al. Survey on deep learning based image super-resolution[J]. Comput Eng Appl, 2021, 57(24): 51-60. DOI: 10.3778/j.issn.1002-8331.2105-0418.
[35]
GAO Y, LIU Z, QIN P L, et al. Medical image super-resolution algorithm based on deep residual generative adversarial network[J]. J Comput Appl, 2018, 38(9): 2689-2695. DOI: 10.11772/j.issn.1001-9081.2018030574.
[36]
CHEN Z L, PAWAR K, EKANAYAKE M, et al. Deep learning for image enhancement and correction in magnetic resonance imaging-state-of-the-art and challenges[J]. J Digit Imaging, 2023, 36(1): 204-230. DOI: 10.1007/s10278-022-00721-9.
[37]
CHEN C X, XIONG L, LIN Y P, et al. Super-resolution reconstruction for early cervical cancer magnetic resonance imaging based on deep learning[J/OL]. Biomed Eng Online, 2024, 23(1): 84 [2025-03-15]. https://pubmed.ncbi.nlm.nih.gov/39175006. DOI: 10.1186/s12938-024-01281-5.

PREV Evaluation of the value of diffusion tensor imaging in assessing the early efficacy of high-intensity focused ultrasound in adenomyosis
NEXT Study on predicting LVSI status in preoperative cervical cancer patients without lymph node metastasis using habitat radiomics based on DCE-MRI quantitative parametric maps
  


LINK


Tel & Fax: +8610-67113815    E-mail: editor@cjmri.cn