Share：
Share this content in WeChat
X
[Chinese][PDF]4995
• Special Focus •
Application of 3D convolutional neural network based on fusion of multiple sequence MRI to evaluate the survival prediction of patients with glioma
YU Xuan  WU Yaping  BAI Yan  WEI Huanhuan  GAO Haiyan  CHEN Lijuan  WANG Meiyun 

Cite this article as: YU X, WU Y P, BAI Y, et al. Application of 3D convolutional neural network based on fusion of multiple sequence MRI to evaluate the survival prediction of patients with glioma[J]. Chin J Magn Reson Imaging, 2023, 14(3): 12-16. DOI:10.12015/issn.1674-8034.2023.03.003.


[Abstract] Objective To explore the survival risk probability of patients with glioma by using 3D convolutional neural network fusion of multiple sequence MRI.Materials and Methods Retrospectively analyzing glioma patients who underwent preoperative cranial MRI examination, 63 patients were included from the picture archiving and communication system of Henan Provincial People's Hospital according to the inclusion and exclusion criterias. T1 weighted image, T2 weighted image, T1 weighted contrast enhanced image, fluidattenuated inversion recovery image data of patients were collected, combined with 500 cases of The Cancer Imaging Archive. Two neuroimaging diagnosticians manually outlined the lesion region of interest on MRIs, divided into high-risk group, medium-risk group and low-risk group according to the patient's survival time. A 3D convolutional neural network deep learning model was constructed, and the dataset was divided into training set, verification set and test set according to 3∶1∶1 to evaluate the survival risk probability of glioma patients.Results The area under the receiver operating characteristic curve (AUC) of the high, medium and low risk groups in the training set was 0.81, 0.79 and 0.86 respectively, the AUC of the high, medium and low risk groups in the validation set was 0.74, 0.78 and 0.81 respectively, and the AUC of the high, medium and low risk groups in the test set was 0.72, 0.74 and 0.75 respectively.Conclusions The deep learning model based on multiple sequence MRI provides auxiliary support for survival prediction of patients with glioma, and supplies quantitative information for doctors in clinical diagnosis and prognosis prediction, which has important scientific value and clinical significance.
[Keywords] glioma；survival prediction；deep learning；3D convolution neural network；magnetic resonance imaging

YU Xuan1   WU Yaping1   BAI Yan1   WEI Huanhuan2   GAO Haiyan1   CHEN Lijuan1   WANG Meiyun1, 3*  

1 Department of Radiology, People's Hospital of Zhengzhou University, Henan Provincial People's Hospital, Zhengzhou 450003, China

2 Academy of Medical Science, Zhengzhou University, Zhengzhou 450000, China

3 Laboratory of Brain Science and Brain-Like Intelligence Technology, Institute for Integrated Medical Science and Engineering, Henan Academy of Sciences, Zhengzhou 450046, China

Corresponding author: Wang MY, E-mail: mywang@zzu.edu.cn

Conflicts of interest   None.

ACKNOWLEDGMENTS Medical Science and Technological Project of Henan Province (No. SBGJ202101002); Natural Science Foundation of Henan Province (No. 212300410240, 222300420354).
Received  2022-12-13
Accepted  2023-03-06
DOI: 10.12015/issn.1674-8034.2023.03.003
Cite this article as: YU X, WU Y P, BAI Y, et al. Application of 3D convolutional neural network based on fusion of multiple sequence MRI to evaluate the survival prediction of patients with glioma[J]. Chin J Magn Reson Imaging, 2023, 14(3): 12-16. DOI:10.12015/issn.1674-8034.2023.03.003.

[1]
WELLER M, WICK W, ALDAPE K, et al. Glioma[J/OL]. Nat Rev Dis Primers, 2015, 1: 15017 [2022-12-13]. https://www.nature.com/articles/nrdp201517. DOI: 10.1038/nrdp.2015.17.
[2]
XU S C, TANG L, LI X Z, et al. Immunotherapy for glioma: current management and future application[J]. Cancer Lett, 2020, 476: 1-12. DOI: 10.1016/j.canlet.2020.02.002.
[3]
WESSELING P, CAPPER D. WHO 2016 classification of gliomas[J]. Neuropathol Appl Neurobiol, 2018, 44(2): 139-150. DOI: 10.1111/nan.12432.
[4]
BAI J, VARGHESE J, JAIN R. Adult glioma WHO classification update, genomics, and imaging: what the radiologists need to know[J]. Top Magn Reson Imaging, 2020, 29(2): 71-82. DOI: 10.1097/RMR.0000000000000234.
[5]
MAIR M J, GEURTS M, VAN DEN BENT M J, et al. A basic review on systemic treatment options in WHO grade Ⅱ-Ⅲ gliomas[J/OL]. Cancer Treat Rev, 2021, 92: 102124 [2022-12-13]. https://www.sciencedirect.com/science/article/pii/S0305737220301626. DOI: 10.1016/j.ctrv.2020.102124.
[6]
DELGADO-LÓPEZ P D, CORRALES-GARCÍA E M, MARTINO J, et al. Diffuse low-grade glioma: a review on the new molecular classification, natural history and current management strategies[J]. Clin Transl Oncol, 2017, 19(8): 931-944. DOI: 10.1007/s12094-017-1631-4.
[7]
YOUSSEF G, MILLER J J. Lower grade gliomas[J/OL]. Curr Neurol Neurosci Rep, 2020, 20(7): 21 [2022-12-13]. https://link.springer.com/content/pdf/10.1007/s11910-020-01040-8.pdf. DOI: 10.1007/s11910-020-01040-8.
[8]
MOLINARO A M, TAYLOR J W, WIENCKE J K, et al. Genetic and molecular epidemiology of adult diffuse glioma[J]. Nat Rev Neurol, 2019, 15(7): 405-417. DOI: 10.1038/s41582-019-0220-2.
[9]
POFF A, KOUTNIK A P, EGAN K M, et al. Targeting the Warburg effect for cancer treatment: Ketogenic diets for management of glioma[J]. Semin Cancer Biol, 2019, 56: 135-148. DOI: 10.1016/j.semcancer.2017.12.011.
[10]
LAPOINTE S, PERRY A, BUTOWSKI N A. Primary brain tumours in adults[J]. Lancet, 2018, 392(10145): 432-446. DOI: 10.1016/S0140-6736(18)30990-5.
[11]
CUI S G, MAO L, JIANG J F, et al. Automatic semantic segmentation of brain gliomas from MRI images using a deep cascaded neural network[J/OL]. J Healthc Eng, 2018, 2018: 4940593 [2022-12-13]. https://pubmed.ncbi.nlm.nih.gov/29755716/. DOI: 10.1155/2018/4940593.
[12]
CHEN H Y, LI C, ZHENG L, et al. A machine learning-based survival prediction model of high grade glioma by integration of clinical and dose-volume histogram parameters[J]. Cancer Med, 2021, 10(8): 2774-2786. DOI: 10.1002/cam4.3838.
[13]
TAN Y, MU W, WANG X C, et al. Improving survival prediction of high-grade glioma via machine learning techniques based on MRI radiomic, genetic and clinical risk factors[J/OL]. Eur J Radiol, 2019, 120: 108609 [2022-12-13]. https://www.sciencedirect.com/science/article/abs/pii/S0720048X19302505. DOI: 10.1016/j.ejrad.2019.07.010.
[14]
MENZE B H, JAKAB A, BAUER S, et al. The multimodal brain tumor image segmentation benchmark (BRATS)[J]. IEEE Trans Med Imaging, 2014, 34(10): 1993-2024. DOI: 10.1109/TMI.2014.2377694.
[15]
BAKAS S, AKBARI H, SOTIRAS A, et al. Advancing The Cancer Genome Atlas glioma MRI collections with expert segmentation labels and radiomic features[J/OL]. Sci Data, 2017, 4: 170117 [2022-12-13]. https://www.nature.com/articles/sdata2017117. DOI: 10.1038/sdata.2017.117.
[16]
WU Y P, LIU B, WU W G, et al. Grading glioma by radiomics with feature selection based on mutual information[J]. J Ambient Intell Human Comput, 2018, 9(5): 1671-1682. DOI: 10.1007/s12652-018-0883-3.
[17]
SMITH S M, JENKINSON M, WOOLRICH M W, et al. Advances in functional and structural MR image analysis and implementation as FSL[J/OL]. NeuroImage, 2004, 23(Suppl 1): S208-S219 [2022-12-13]. https://www.sciencedirect.com/science/article/abs/pii/S1053811904003933. DOI: 10.1016/j.neuroimage.2004.07.051.
[18]
VICKERY S, HOPKINS W D, SHERWOOD C C, et al. Chimpanzee brain morphometry utilizing standardized MRI preprocessing and macroanatomical annotations[J/OL]. Elife, 2020, 9: e60136 [2022-12-13]. https://pubmed.ncbi.nlm.nih.gov/33226338/. DOI: 10.7554/eLife.60136.
[19]
VAKANSKI A, XIAN M, FREER P E. Attention-enriched deep learning model for breast tumor segmentation in ultrasound images[J]. Ultrasound Med Biol, 2020, 46(10): 2819-2833. DOI: 10.1016/j.ultrasmedbio.2020.06.015.
[20]
YU X, WU Y P, BAI Y, et al. A lightweight 3D UNet model for glioma grading[J/OL]. Phys Med Biol, 2022, 67(15) [2022-12-13]. https://iopscience.iop.org/article/10.1088/1361-6560/ac7d33. DOI: 10.1088/1361-6560/ac7d33.
[21]
SZENTIMREY Z, DE RIBAUPIERRE S, FENSTER A, et al. Automated 3D U-net based segmentation of neonatal cerebral ventricles from 3D ultrasound images[J]. Med Phys, 2022, 49(2): 1034-1046. DOI: 10.1002/mp.15432.
[22]
SAGBERG L M, JAKOLA A S, REINERTSEN I, et al. How well do neurosurgeons predict survival in patients with high-grade glioma?[J]. Neurosurg Rev, 2022, 45(1): 865-872. DOI: 10.1007/s10143-021-01613-2.
[23]
NICHOLSON J G, FINE H A. Diffuse glioma heterogeneity and its therapeutic implications[J]. Cancer Discov, 2021, 11(3): 575-590. DOI: 10.1158/2159-8290.CD-20-1474.
[24]
COUPET M, URRUTY T, LEELANUPAB T, et al. A multi-sequences MRI deep framework study applied to glioma classfication[J]. Multimed Tools Appl, 2022, 81(10): 13563-13591. DOI: 10.1007/s11042-022-12316-1.
[25]
BI W L, HOSNY A, SCHABATH M B, et al. Artificial intelligence in cancer imaging: clinical challenges and applications[J]. CA Cancer J Clin, 2019, 69(2): 127-157. DOI: 10.3322/caac.21552.
[26]
JIN W N, FATEHI M, ABHISHEK K, et al. Artificial intelligence in glioma imaging: challenges and advances[J/OL]. J Neural Eng, 2020, 17(2): 021002 [2022-12-13]. https://pubmed.ncbi.nlm.nih.gov/32191935/. DOI: 10.1088/1741-2552/ab8131.
[27]
ZHAO H, BAI Y, WANG M Y. Progress of multimodality magnetic resonance imaging in genotyping and prognostic evaluation of gliomas[J]. Chin J Magn Reson Imaging, 2021, 12(9): 98-102. DOI: 10.12015/issn.1674-8034.2021.09.025.
[28]
ZHAO W W, SUN J, ZHU J Q. Research progress of artificial intelligence in MRI diagnosis of glioma[J]. Chin J Magn Reson Imaging, 2021, 12(8): 88-90. DOI: 10.12015/issn.1674-8034.2021.08.019.
[29]
ZHUGE Y, NING H, MATHEN P, et al. Automated glioma grading on conventional MRI images using deep convolutional neural networks[J]. Med Phys, 2020, 47(7): 3044-3053. DOI: 10.1002/mp.14168.
[30]
ALI M J, RAZA B, SHAHID A R. Multi-level kronecker convolutional neural network (ML-KCNN) for glioma segmentation from multi-modal MRI volumetric data[J]. J Digit Imaging, 2021, 34(4): 905-921. DOI: 10.1007/s10278-021-00486-7.
[31]
MLADENOVSK M, VALKOV I, OVCHAROV M, et al. High grade glioma surgery-clinical aspects and prognosis[J]. Folia Med (Plovdiv), 2021, 63(1): 35-41. DOI: 10.3897/folmed.63.e55255.
[32]
SCOTT J N, BRASHER P A, SEVICK R J, et al. How often are nonenhancing supratentorial gliomas malignant? A population study[J]. Neurology, 2002, 59(6): 947-949. DOI: 10.1212/wnl.59.6.947.
[33]
WEN P Y, MACDONALD D R, REARDON D A, et al. Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology working group[J]. J Clin Oncol, 2010, 28(11): 1963-1972. DOI: 10.1200/JCO.2009.26.3541.

PREV Prediction of mixed ischemic stroke mechanism based on HR-MRI radiomics of intracranial arterial plaque
NEXT 3D-ultrashort echo time MRI-based radiomics model facilitates the assessment of lymph node metastasis in non-small cell lung cancer
  


LINK


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