Share：
Share this content in WeChat
X
[Chinese][PDF]1174
• Clinical Article •
Study on the ability to grade the risk of cervical spondylotic myelopathy by using machine learning model based on MRI radiomics
XU Gang  CHEN Peng  LI Yulong  ZHU Yun  XIE Zongyu 

Cite this article as: XU G, CHEN P, LI Y L, et al. Study on the ability to grade the risk of cervical spondylotic myelopathy by using machine learning model based on MRI radiomics[J]. Chin J Magn Reson Imaging, 2024, 15(4): 50-55, 77. DOI:10.12015/issn.1674-8034.2024.04.009.


[Abstract] Objective To explore the value of machine learning (ML) model based on MRI radiomics features in grading the risk of cervical spondylotic myelopathy (CSM).Materials and Methods This retrospective study included 317 patients diagnosed with cervical spondylotic myelopathy (CSM), according to the Japanese Orthopaedic Association (JOA) score they were divided into mild CSM group (193 patients) and moderate-severe CSM group (124 patients). Spinal cord in the transverse T2-weighted MR images were manually sketched to generate a region of interest (ROI) and extract radiomics features. The Z-Score standardization were used to unify metrics. The Pearson correlation coefficients (PCC) and recursive feature elimination (RFE) were used to reduce the dimension and select the feature. Various ML algorithms including logistic regression (LR), adaboost (AB), native bayes (NB), support vector machine (SVM) were used to build ML models. The area under the curve (AUC) of receiver operating characteristic were used to evaluate the diagnostic efficacy of the model.Results A total of 15 radiomics salient features were selected to build models, SVM (training set AUC vs. test set AUC: 0.833 vs. 0.813) and LR (0.831 vs. 0.812) have good grading ability and are more stable among the four classifiers, there were no statistically significant differences between the models. AB classifier has the best grading ability in the training group (AUC=0.984) but poor grading ability in the test group (AUC=0.725), The AB classifier model has lower stabilization than SVM and LR classifier model.Conclusions ML model based on MRI radiomics has good risk grading ability for CSM, which can provide certain reference value for clinical preoperative diagnosis.
[Keywords] cervical spondylotic myelopathy；radiomics；machine learning；risk classification；magnetic resonance imaging

XU Gang1   CHEN Peng2   LI Yulong3   ZHU Yun4   XIE Zongyu4*  

1 Department of Medical Imaging, Anhui University of Science & Technology Affiliated Huainan Xinhua Hospital, Huainan 232000, China

2 Department of Radiology, Huzhou Central Hospital, Huzhou 313000, China

3 Department of Spinal Surgery, Anhui University of Science & Technology Affiliated Huainan Xinhua Hospital, Huainan 232000, China

4 Department of Radiology, The First Affiliated Hospital of Bengbu Medical University, Bengbu 233000, China

Corresponding author: XIE Z Y, E-mail: zongyuxie@sina.com

Conflicts of interest   None.

Received  2023-10-17
Accepted  2024-03-22
DOI: 10.12015/issn.1674-8034.2024.04.009
Cite this article as: XU G, CHEN P, LI Y L, et al. Study on the ability to grade the risk of cervical spondylotic myelopathy by using machine learning model based on MRI radiomics[J]. Chin J Magn Reson Imaging, 2024, 15(4): 50-55, 77. DOI:10.12015/issn.1674-8034.2024.04.009.

[1]
IYER A, AZAD T D, THARIN S. Cervical spondylotic myelopathy[J]. Clin Spine Surg, 2016, 29(10): 408-414. DOI: 10.1097/BSD.0000000000000397.
[2]
BAKHSHESHIAN J, MEHTA V A, LIU J C. Current diagnosis and management of cervical spondylotic myelopathy[J]. Global Spine J, 2017, 7(6): 572-586. DOI: 10.1177/2192568217699208.
[3]
MCCORMICK J R, SAMA A J, SCHILLER N C, et al. Cervical spondylotic myelopathy: a guide to diagnosis and management[J]. J Am Board Fam Med, 2020, 33(2): 303-313. DOI: 10.3122/jabfm.2020.02.190195.
[4]
MAKHCHOUNE M, TRIFFAUX M, BOURAS T, et al. The value of dynamic MRI in cervical spondylotic myelopathy: about 24 cases[J/OL]. Ann Med Surg (Lond), 2022, 83: 104717 [2023-07-26]. https://pubmed.ncbi.nlm.nih.gov/36389194. DOI: 10.1016/j.amsu.2022.104717.
[5]
XU Z A, XIAO L, LIU C, et al. Correlation analysis of surgical efficacy and risk factors of cervical spondylotic myelopathy with high signal intensity on MRI-T2WI[J]. Curr Med Imaging, 2023, 19(2): 142-148. DOI: 10.2174/1573405618666220111121650.
[6]
HE B, SHELDRICK K, DAS A, et al. Clinical and research MRI techniques for assessing spinal cord integrity in degenerative cervical myelopathy-a scoping review[J/OL]. Biomedicines, 2022, 10(10): 2621 [2023-07-26]. https://pubmed.ncbi.nlm.nih.gov/36289883. DOI: 10.3390/biomedicines10102621.
[7]
MARTIN A R, ALEKSANDEREK I, COHEN-ADAD J, et al. Translating state-of-the-art spinal cord MRI techniques to clinical use: a systematic review of clinical studies utilizing DTI, MT, MWF, MRS, and fMRI[J]. Neuroimage Clin, 2016, 10: 192-238. DOI: 10.1016/j.nicl.2015.11.019.
[8]
LAMBIN P, RIOS-VELAZQUEZ E, LEIJENAAR R, et al. Radiomics: extracting more information from medical images using advanced feature analysis[J]. Eur J Cancer, 2012, 48(4): 441-446. DOI: 10.1016/j.ejca.2011.11.036.
[9]
RAJKOMAR A, DEAN J, KOHANE I. Machine learning in medicine[J]. N Engl J Med, 2019, 380(14): 1347-1358. DOI: 10.1056/nejmra1814259.
[10]
CHOI R Y, COYNER A S, KALPATHY-CRAMER J, et al. Introduction to machine learning, neural networks, and deep learning[J/OL]. Transl Vis Sci Technol, 2020, 9(2): 14 [2023-07-15]. https://pubmed.ncbi.nlm.nih.gov/32704420. DOI: 10.1167/tvst.9.2.14.
[11]
ZHANG M Z, OU-YANG H Q, LIU J F, et al. Predicting postoperative recovery in cervical spondylotic myelopathy: construction and interpretation of T2*-weighted radiomic-based extra trees models[J]. Eur Radiol, 2022, 32(5): 3565-3575. DOI: 10.1007/s00330-021-08383-x.
[12]
HOPKINS B S, WEBER K A, KESAVABHOTLA K, et al. Machine learning for the prediction of cervical spondylotic myelopathy: a post hoc pilot study of 28 participants[J/OL]. World Neurosurg, 2019, 127: e436-e442 [2023-07-26]. https://www.sciencedirect.com/science/article/abs/pii/S1878875019308459. DOI: 10.1016/j.wneu.2019.03.165.
[13]
LEE G W, SHIN H, CHANG M C. Deep learning algorithm to evaluate cervical spondylotic myelopathy using lateral cervical spine radiograph[J/OL]. BMC Neurol, 2022, 22(1): 147 [2023-07-26]. https://pubmed.ncbi.nlm.nih.gov/35443618. DOI: 10.1186/s12883-022-02670-w.
[14]
ALBARADEI S, THAFAR M, ALSAEDI A, et al. Machine learning and deep learning methods that use omics data for metastasis prediction[J]. Comput Struct Biotechnol J, 2021, 19: 5008-5018. DOI: 10.1016/j.csbj.2021.09.001.
[15]
WANG S Q, LI X, CUI J L, et al. Prediction of myelopathic level in cervical spondylotic myelopathy using diffusion tensor imaging[J]. J Magn Reson Imaging, 2015, 41(6): 1682-1688. DOI: 10.1002/jmri.24709.
[16]
ZHANG M Z, OU-YANG H Q, JIANG L, et al. Optimal machine learning methods for radiomic prediction models: clinical application for preoperative T2*-weighted images of cervical spondylotic myelopathy[J/OL]. JOR Spine, 2021, 4(4): e1178 [2023-07-26]. https://pubmed.ncbi.nlm.nih.gov/35005444. DOI: 10.1002/jsp2.1178.
[17]
WATANABE M, CHIKUDA H, FUJIWARA Y, et al. Japanese Orthopaedic Association (JOA) Clinical practice guidelines on the Management of Cervical Spondylotic Myelopathy, 2020-Secondary publication[J]. J Orthop Sci, 2023, 28(1): 1-45. DOI: 10.1016/j.jos.2022.03.012.
[18]
GBD MORTALITY AND CAUSES OF DEATH COLLABORATORS. Global, regional, and national age-sex specific all-cause and cause-specific mortality for 240 causes of death, 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013[J]. Lancet, 2015, 385(9963): 117-171. DOI: 10.1016/S0140-6736(14)61682-2.
[19]
AMES C P, BLONDEL B, SCHEER J K, et al. Cervical radiographical alignment: comprehensive assessment techniques and potential importance in cervical myelopathy[J]. Spine, 2013, 38(22Suppl 1): S149-S160. DOI: 10.1097/BRS.0b013e3182a7f449.
[20]
MELANCIA J L, FRANCISCO A F, ANTUNES J L. Spinal stenosis[J]. Handbook of clinical neurology, 2014, 119: 541-549. DOI: 10.1016/B978-0-7020-4086-3.00035-7.
[21]
HOPKINS B S, WEBER K A, CLONEY M B, et al. Tract-specific volume loss on 3T MRI in patients with cervical spondylotic myelopathy[J/OL]. Spine, 2018, 43(20): E1204-E1209 [2023-07-26]. https://pubmed.ncbi.nlm.nih.gov/29649085. DOI: 10.1097/BRS.0000000000002667.
[22]
YOU J Y, LEE J W, LEE E, et al. MR classification system based on axial images for cervical compressive myelopathy[J]. Radiology, 2015, 276(2): 553-561. DOI: 10.1148/radiol.2015142384.
[23]
KIM T H, HA Y, SHIN J J, et al. Signal intensity ratio on magnetic resonance imaging as a prognostic factor in patients with cervical compressive myelopathy[J/OL]. Medicine, 2016, 95(39): e4649 [2023-07-26]. https://pubmed.ncbi.nlm.nih.gov/27684796. DOI: 10.1097/MD.0000000000004649.
[24]
ZHAO W L, ZHANG E L, LIU K, et al. Development of artificial intelligence in diagnosis and treatment of spinal diseases[J]. Chin J Magn Reson Imag, 2022, 13(6): 160-163. DOI: 10.12015/issn.1674-8034.2022.06.034.
[25]
XU S J, ZHAO P, CHEN X M, et al. Diffusion tensor imaging of the lumbosacral enlargement helps to estimate the neurologic recovery after decompression surgery in patients with cervical spondylotic myelpoathy[J]. Chin J Magn Reson Imag, 2022, 13(6): 98-101, 107. DOI: 10.12015/issn.1674-8034.2022.06.019.
[26]
MERALI Z, WANG J Z, BADHIWALA J H, et al. A deep learning model for detection of cervical spinal cord compression in MRI scans[J/OL]. Sci Rep, 2021, 11(1): 10473 [2023-07-26]. https://www.nature.com/articles/s41598-021-89848-3. DOI: 10.1038/s41598-021-89848-3.
[27]
PRIPP A H. Pearson's or Spearman's correlation coefficients[J/OL]. Tidsskr Nor Laegeforen, 2018, 138(8) [2023-07-26]. https://pubmed.ncbi.nlm.nih.gov/29737766. DOI: 10.4045/tidsskr.18.0042.
[28]
HUMPHREYS R K, PUTH M T, NEUHÄUSER M, et al. Underestimation of Pearson's product moment correlation statistic[J]. Oecologia, 2019, 189(1): 1-7. DOI: 10.1007/s00442-018-4233-0.
[29]
JHA S C, MIYAZAKI M, TSUMURA H. Kinetic change of spinal cord compression on flexion-extension magnetic resonance imaging in cervical spine[J]. Clin Neurol Neurosurg, 2018, 174: 86-91. DOI: 10.1016/j.clineuro.2018.09.017.
[30]
OMORI M, SHIBUYA S, NAKAJIMA T, et al. Hand dexterity impairment in patients with cervical myelopathy: a new quantitative assessment using a natural prehension movement[J/OL]. Behav Neurol, 2018, 2018: 5138234 [2023-07-36]. https://pubmed.ncbi.nlm.nih.gov/30073036. DOI: 10.1155/2018/5138234.
[31]
HOLLY L T, FREITAS B, MCARTHUR D L, et al. Proton magnetic resonance spectroscopy to evaluate spinal cord axonal injury in cervical spondylotic myelopathy[J]. J Neurosurg Spine, 2009, 10(3): 194-200. DOI: 10.3171/2008.12.SPINE08367.
[32]
HUANG S J, CAI N G, PACHECO P P, et al. Applications of support vector machine (SVM) learning in cancer genomics[J]. Cancer Genomics Proteomics, 2018, 15(1): 41-51. DOI: 10.21873/cgp.20063.
[33]
BUCHLAK Q D, ESMAILI N, LEVEQUE J C, et al. Machine learning applications to clinical decision support in neurosurgery: an artificial intelligence augmented systematic review[J]. Neurosurg Rev, 2020, 43(5): 1235-1253. DOI: 10.1007/s10143-019-01163-8.
[34]
SEVINÇ E. An empowered AdaBoost algorithm implementation: a COVID-19 dataset study[J/OL]. Comput Ind Eng, 2022, 165: 107912 [2023-07-26]. https://www.sciencedirect.com/science/article/pii/S0360835221008160. DOI: 10.1016/j.cie.2021.107912.
[35]
BOURS M J. Bayes' rule in diagnosis[J]. J Clin Epidemiol, 2021, 131: 158-160. DOI: 10.1016/j.jclinepi.2020.12.021.

PREV Value of three-dimensional arterial spin labeling in distinguishing between benign and malignant thyroid nodules
NEXT Study on quantitative assessment of left ventricular dysfunction in patients with COPD by cardiovascular magnetic resonance feature tracking
  


LINK


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