Journal of Neuro-Oncology

, Volume 139, Issue 2, pp 491–499 | Cite as

Machine learning: a useful radiological adjunct in determination of a newly diagnosed glioma’s grade and IDH status

  • Céline De Looze
  • Alan Beausang
  • Jane Cryan
  • Teresa Loftus
  • Patrick G. Buckley
  • Michael Farrell
  • Seamus Looby
  • Richard Reilly
  • Francesca Brett
  • Hugh Kearney
Clinical Study



Machine learning methods have been introduced as a computer aided diagnostic tool, with applications to glioma characterisation on MRI. Such an algorithmic approach may provide a useful adjunct for a rapid and accurate diagnosis of a glioma. The aim of this study is to devise a machine learning algorithm that may be used by radiologists in routine practice to aid diagnosis of both: WHO grade and IDH mutation status in de novo gliomas.


To evaluate the status quo, we interrogated the accuracy of neuroradiology reports in relation to WHO grade: grade II 96.49% (95% confidence intervals [CI] 0.88, 0.99); III 36.51% (95% CI 0.24, 0.50); IV 72.9% (95% CI 0.67, 0.78). We derived five MRI parameters from the same diagnostic brain scans, in under two minutes per case, and then supplied these data to a random forest algorithm.


Machine learning resulted in a high level of accuracy in prediction of tumour grade: grade II/III; area under the receiver operating characteristic curve (AUC) = 98%, sensitivity = 0.82, specificity = 0.94; grade II/IV; AUC = 100%, sensitivity = 1.0, specificity = 1.0; grade III/IV; AUC = 97%, sensitivity = 0.83, specificity = 0.97. Furthermore, machine learning also facilitated the discrimination of IDH status: AUC of 88%, sensitivity = 0.81, specificity = 0.77.


These data demonstrate the ability of machine learning to accurately classify diffuse gliomas by both WHO grade and IDH status from routine MRI alone—without significant image processing, which may facilitate usage as a diagnostic adjunct in clinical practice.


Diagnostic accuracy Machine learning Glioma Random forest MRI 


Compliance with ethical standards

Conflict of interest

None of the authors of this study have any conflict of interest in relation to this work.

Human and animal participants

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent

Informed consent was obtained from all participants in this study.


  1. 1.
    Ostrom QT, Gittleman H, Fulop J et al (2015) CBTRUS statistical report: primary brain and central nervous system tumors diagnosed in the United States in 2008–2012. Neuro Oncol 17:iv1–iv62. CrossRefPubMedPubMedCentralGoogle Scholar
  2. 2.
    Gittleman H, Kromer C, Ostrom QT et al (2017) Is mortality due to primary malignant brain and other central nervous system tumors decreasing? J Neurooncol 133:265–275. CrossRefPubMedGoogle Scholar
  3. 3.
    Louis DN, Perry A, Reifenberger G et al (2016) The 2016 World Health Organization classification of tumors of the central nervous system: a summary. Acta Neuropathol 131:803–820. CrossRefPubMedGoogle Scholar
  4. 4.
    Yan H, Parsons DW, Jin G et al (2009) IDH1 and IDH2 Mutations in Gliomas. N Engl J Med 360:765–773. CrossRefPubMedPubMedCentralGoogle Scholar
  5. 5.
    Metellus P, Coulibaly B, Colin C et al (2010) Absence of IDH mutation identifies a novel radiologic and molecular subtype of WHO grade II gliomas with dismal prognosis. Acta Neuropathol 120:719–729. CrossRefPubMedGoogle Scholar
  6. 6.
    Thrall JH, Li X, Li Q et al (2018) Artificial intelligence and machine learning in radiology: opportunities, challenges, pitfalls, and criteria for success. J Am Coll Radiol 15:504–508. CrossRefPubMedGoogle Scholar
  7. 7.
    Kohli M, Prevedello LM, Filice RW, Geis JR (2017) Implementing machine learning in radiology practice and research. Am J Roentgenol 208:754–760CrossRefGoogle Scholar
  8. 8.
    Zacharaki EI, Wang S, Chawla S, Soo D (2010) Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme. Magn Reson Med 62:1609–1618. CrossRefGoogle Scholar
  9. 9.
    Zacharaki EI, Morita N, Bhatt P et al (2012) Survival analysis of patients with high-grade gliomas based on data mining of imaging variables. AJNR Am J Neuroradiol 33:1065–1071. CrossRefPubMedPubMedCentralGoogle Scholar
  10. 10.
    Svolos P, Tsolaki E, Kapsalaki E et al (2013) Investigating brain tumor differentiation with diffusion and perfusion metrics at 3T MRI using pattern recognition techniques. Magn Reson Imaging 31:1567–1577. CrossRefPubMedGoogle Scholar
  11. 11.
    Wen PY, Reardon DA (2016) Neuro-oncology in 2015: progress in glioma diagnosis, classification and treatment. Nat Rev Neurol 12:2015–2016. CrossRefGoogle Scholar
  12. 12.
    Kondziolka D, Lunsford LD, Martinez a J (1993) Unreliability of contemporary neurodiagnostic imaging in evaluating suspected adult supratentorial (low-grade) astrocytoma. J Neurosurg 79:533–536. CrossRefPubMedGoogle Scholar
  13. 13.
    Aronen HJ, Gazit IE, Louis DN et al (1994) Cerebral blood volume maps of gliomas: comparison with tumor grade and histologic findings. Radiology 191:41–51. CrossRefPubMedGoogle Scholar
  14. 14.
    Knopp EA, Cha S, Johnson G et al (1999) Glial neoplasms: dynamic contrast-enhanced T2*-weighted MR imaging. Radiology 211:791–798. CrossRefPubMedGoogle Scholar
  15. 15.
    Guzman-De-Villoria JA, Mateos-Perez JM, Fernandez-Garcia P et al (2014) Added value of advanced over conventional magnetic resonance imaging in grading gliomas and other primary brain tumors. Cancer Imaging 14:1–10. CrossRefGoogle Scholar
  16. 16.
    Miloushev VZ, Chow DS, Filippi CG (2015) Meta-analysis of diffusion metrics for the prediction of tumor grade in gliomas. AJNR Am J Neuroradiol 36:302–308. CrossRefPubMedGoogle Scholar
  17. 17.
    Ranjith G, Parvathy R, Vikas V et al (2015) Machine learning methods for the classification of gliomas: initial results using features extracted from MR spectroscopy. Neuroradiol J 28:106–111. CrossRefPubMedPubMedCentralGoogle Scholar
  18. 18.
    Wiestler B, Kluge A, Lukas M et al (2016) Multiparametric MRI-based differentiation of WHO grade II/III glioma and WHO grade IV glioblastoma. Sci Rep 6:35142. CrossRefPubMedPubMedCentralGoogle Scholar
  19. 19.
    Flavahan WA, Drier Y, Liau BB et al (2016) Insulator dysfunction and oncogene activation in IDH mutant gliomas. Nature 529:110–114. CrossRefPubMedGoogle Scholar
  20. 20.
    Sanson M, Marie Y, Paris S et al (2009) Isocitrate dehydrogenase 1 codon 132 mutation is an important prognostic biomarker in gliomas. J Clin Oncol 27:4150–4154. CrossRefPubMedGoogle Scholar
  21. 21.
    Beiko J, Suki D, Hess KR et al (2014) IDH1 mutant malignant astrocytomas are more amenable to surgical resection and have a survival benefit associated with maximal surgical resection. Neuro Oncol 16:81–91. CrossRefPubMedGoogle Scholar
  22. 22.
    Qi S, Yu L, Li H et al (2014) Isocitrate dehydrogenase mutation is associated with tumor location and magnetic resonance imaging characteristics in astrocytic neoplasms. Oncol Lett 7:1895–1902. CrossRefPubMedPubMedCentralGoogle Scholar
  23. 23.
    Delfanti RL, Piccioni DE, Handwerker J et al (2017) Imaging correlates for the 2016 update on WHO classification of grade II/III gliomas: implications for IDH, 1p/19q and ATRX status. J Neurooncol 135:601–609. CrossRefPubMedGoogle Scholar
  24. 24.
    Branzoli F, Di Stefano AL, Capelle L et al (2017) Highly specific determination of IDH status using edited in vivo magnetic resonance spectroscopy. Neuro Oncol. CrossRefGoogle Scholar
  25. 25.
    Zhou H, Vallières M, Bai HX et al (2017) MRI features predict survival and molecular markers in diffuse lower-grade gliomas. Neuro Oncol 19:862–870. CrossRefPubMedPubMedCentralGoogle Scholar
  26. 26.
    Zhang B, Chang K, Ramkissoon S et al (2017) Multimodal MRI features predict isocitrate dehydrogenase genotype in high-grade gliomas. Neuro Oncol 19:109–117. CrossRefPubMedGoogle Scholar
  27. 27.
    Elkhaled A, Jalbert LE, Phillips JJ et al (2012) Magnetic resonance of 2-hydroxyglutarate in IDH1-mutated low-grade gliomas. Sci Transl Med 4:116ra5–116ra5. CrossRefPubMedPubMedCentralGoogle Scholar
  28. 28.
    Reuss DE, Sahm F, Schrimpf D et al (2015) ATRX and IDH1-R132H immunohistochemistry with subsequent copy number analysis and IDH sequencing as a basis for an “integrated” diagnostic approach for adult astrocytoma, oligodendroglioma and glioblastoma. Acta Neuropathol 129:133–146. CrossRefPubMedGoogle Scholar
  29. 29.
    Arita H, Narita Y, Matsushita Y et al (2015) Development of a robust and sensitive pyrosequencing assay for the detection of IDH1/2 mutations in gliomas. Brain Tumor Pathol 32:22–30. CrossRefPubMedGoogle Scholar
  30. 30.
    Bakas S, Akbari H, Sotiras A et al (2017) Advancing the cancer genome atlas glioma MRI collections with expert segmentation labels and radiomic features. Sci data 4:170117. CrossRefPubMedPubMedCentralGoogle Scholar
  31. 31.
    Clark K, Vendt B, Smith K et al (2013) The Cancer imaging archive (TCIA): maintaining and operating a public information repository. J Digit Imaging 26:1045–1057. CrossRefPubMedPubMedCentralGoogle Scholar
  32. 32.
    Smits M, van den Bent MJ (2017) Imaging correlates of adult glioma genotypes. Radiology 284:316–331. CrossRefPubMedGoogle Scholar
  33. 33.
    Lee EJ, terBrugge K, Mikulis D et al (2011) Diagnostic value of peritumoral minimum apparent diffusion coefficient for differentiation of glioblastoma multiforme from solitary metastatic lesions. Am J Roentgenol 196:71–76. CrossRefGoogle Scholar
  34. 34.
    Breiman L (2001) Statistical modeling: the two cultures (with comments and a rejoinder by the author). Stat Sci 16:199–231. CrossRefGoogle Scholar
  35. 35.
    Zikic D, Glocker B, Konukoglu E et al (2012) Decision forests for tissue-specific segmentation of high-grade gliomas in multi-channel MR. Med Image Comput Comput Assist Interv 15:369–376PubMedGoogle Scholar
  36. 36.
    Fernández-Delgado M, Cernadas E, Barro S et al (2014) Do we need hundreds of classifiers to solve real world classification problems? J Mach Learn Res 15:3133–3181Google Scholar
  37. 37.
    Bland JM, Altman DG (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307–310CrossRefPubMedGoogle Scholar
  38. 38.
    Srivastava N, Hinton G, Krizhevsky A et al (2014) Dropout: a simple way to prevent neural networks from overfitting. J Mach Learn Res 15:1929–1958Google Scholar
  39. 39.
    Chawla NV, Bowyer KW, Hall LO, Kegelmeyer WP (2002) SMOTE: synthetic minority over-sampling technique. J Artif Intell Res 16:321–357. CrossRefGoogle Scholar
  40. 40.
    Williams LH, Drew T (2017) Distraction in diagnostic radiology: how is search through volumetric medical images affected by interruptions? Cogn Res Princ Implic 2:12. CrossRefPubMedPubMedCentralGoogle Scholar
  41. 41.
    Waite S, Kolla S, Jeudy J et al (2017) Tired in the reading room: the influence of fatigue in radiology. J Am Coll Radiol 14:191–197. CrossRefPubMedGoogle Scholar
  42. 42.
    Lee CS, Nagy PG, Weaver SJ, Newman-Toker DE (2013) Cognitive and system factors contributing to diagnostic errors in radiology. Am J Roentgenol 201:611–617CrossRefGoogle Scholar
  43. 43.
    Emblem KE, Pinho MC, Zöllner FG et al (2015) A generic support vector machine model for preoperative glioma survival associations. Radiology 275:228–234. CrossRefPubMedGoogle Scholar
  44. 44.
    Kickingereder P, Bonekamp D, Nowosielski M et al (2016) Radiogenomics of Glioblastoma: Machine Learning-based Classification of Molecular Characteristics by Using Multiparametric and Multiregional MR Imaging Features. Radiology 281:907–918. CrossRefPubMedGoogle Scholar
  45. 45.
    Zacharaki EI, Wang S, Chawla S et al (2009) Classification of brain tumor type and grade using MRI texture and shape in a machine learning scheme. Magn Reson Med 62:1609–1618. CrossRefPubMedPubMedCentralGoogle Scholar
  46. 46.
    Halevy A, Norvig P, Pereira F (2009) The Unreasonable effectiveness of data. IEEE Intell Syst 24:8–12. CrossRefGoogle Scholar
  47. 47.
    Wang Y, Zhang T, Li S et al (2015) Anatomical localization of isocitrate dehydrogenase 1 mutation: a voxel-based radiographic study of 146 low-grade gliomas. Eur J Neurol 22:348–354. CrossRefPubMedGoogle Scholar
  48. 48.
    Horská A, Barker PB (2010) Imaging of brain tumors: MR spectroscopy and metabolic imaging. Neuroimaging Clin N Am 20:293–310. CrossRefPubMedPubMedCentralGoogle Scholar
  49. 49.
    Jain R, Poisson LM, Gutman D et al (2014) Outcome prediction in patients with glioblastoma by using imaging, clinical, and genomic biomarkers: focus on the nonenhancing component of the tumor. Radiology 272:484–493. CrossRefPubMedPubMedCentralGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  • Céline De Looze
    • 1
    • 2
  • Alan Beausang
    • 3
  • Jane Cryan
    • 3
  • Teresa Loftus
    • 4
  • Patrick G. Buckley
    • 4
    • 5
  • Michael Farrell
    • 3
  • Seamus Looby
    • 6
  • Richard Reilly
    • 1
    • 7
    • 8
  • Francesca Brett
    • 3
  • Hugh Kearney
    • 3
  1. 1.Trinity Centre for BioengineeringTrinity College DublinDublinIreland
  2. 2.School of EngineeringTrinity College DublinDublinIreland
  3. 3.Department of NeuropathologyBeaumont HospitalDublinIreland
  4. 4.Department of Molecular PathologyBeaumont HospitalDublinIreland
  5. 5.Genomics Medicine IrelandDublinIreland
  6. 6.Department of NeuroradiologyBeaumont HospitalDublinIreland
  7. 7.Institute of NeurosciencesTrinity College DublinDublinIreland
  8. 8.School of MedicineTrinity College DublinDublinIreland

Personalised recommendations