Predicting response to somatostatin analogues in acromegaly: machine learning-based high-dimensional quantitative texture analysis on T2-weighted MRI

  • Burak KocakEmail author
  • Emine Sebnem Durmaz
  • Pinar Kadioglu
  • Ozge Polat Korkmaz
  • Nil Comunoglu
  • Necmettin Tanriover
  • Naci Kocer
  • Civan Islak
  • Osman Kizilkilic



To investigate the value of machine learning (ML)-based high-dimensional quantitative texture analysis (qTA) on T2-weighted magnetic resonance imaging (MRI) in predicting response to somatostatin analogues (SA) in acromegaly patients with growth hormone (GH)-secreting pituitary macroadenoma, and to compare the qTA with quantitative and qualitative T2-weighted relative signal intensity (rSI) and immunohistochemical evaluation.


Forty-seven patients (24 responsive; 23 resistant patients to SA) were eligible for this retrospective study. Coronal T2-weighted images were used for qTA and rSI evaluation. The immunohistochemical evaluation was based on the granulation pattern of the adenomas. Dimension reduction was carried out by reproducibility analysis and wrapper-based algorithm. ML classifiers were k-nearest neighbours (k-NN) and C4.5 algorithm. The reference standard was the biochemical response status. Predictive performance of qTA was compared with those of the quantitative and qualitative rSI and immunohistochemical evaluation.


Five hundred thirty-five out of 828 texture features had excellent reproducibility. For the qTA, k-NN correctly classified 85.1% of the macroadenomas regarding response to SAs with an area under the receiver operating characteristic curve (AUC-ROC) of 0.847. The accuracy and AUC-ROC ranges of the other methods were 57.4–70.2% and 0.575–0.704, respectively. Differences in predictive performance between qTA-based classification and the other methods were significant (p < 0.05).


The ML-based qTA of T2-weighted MRI is a potential non-invasive tool in predicting response to SAs in patients with acromegaly and GH-secreting pituitary macroadenoma. The method performed better than the qualitative and quantitative rSI and immunohistochemical evaluation.

Key Points

• Machine learning-based texture analysis of T2-weighted MRI can correctly classify response to somatostatin analogues in more than four fifths of the patients.

• Machine learning-based texture analysis performs better than qualitative and quantitative evaluation of relative T2 signal intensity and immunohistochemical evaluation.

• About one third of the texture features may not be excellently reproducible, indicating that a reliability analysis is necessary before model development.


Acromegaly Growth hormone-secreting pituitary adenoma Machine learning Magnetic resonance imaging Somatostatin 







Area under the receiver operating characteristic curve


Growth hormone


Intra-class correlation coefficient


Insulin-like growth factor-1


k-nearest neighbours


Laplacian of Gaussian


Machine learning


Magnetic resonance imaging


Quantitative texture analysis


Region of interest


Relative signal intensity


Somatostatin analogue


Standard deviation


Waikato Environment for Knowledge Analysis



The authors state that this work has not received any funding.

Compliance with ethical standards


The scientific guarantor of this publication is Burak Kocak, MD.

Conflict of interest

The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.

Statistics and biometry

One of the authors (Burak Kocak, MD) has significant statistical expertise.

Informed consent

Written informed consent was waived by the Institutional Review Board.

Ethical approval

Institutional Review Board approval was obtained.


• retrospective

• diagnostic or prognostic study

• performed at one institution

Supplementary material

330_2018_5876_MOESM1_ESM.docx (20 kb)
ESM 1 (DOCX 20 kb)


  1. 1.
    Paragliola RM, Corsello SM, Salvatori R (2017) Somatostatin receptor ligands in acromegaly: clinical response and factors predicting resistance. Pituitary 20:109–115. CrossRefPubMedGoogle Scholar
  2. 2.
    Melmed S, Bronstein MD, Chanson P et al (2018) A consensus statement on acromegaly therapeutic outcomes. Nat Rev Endocrinol 14:552–561. CrossRefPubMedGoogle Scholar
  3. 3.
    Bonneville JF, Bonneville F, Cattin F (2005) Magnetic resonance imaging of pituitary adenomas. Eur Radiol 15:543–548. CrossRefPubMedGoogle Scholar
  4. 4.
    Puig-Domingo M, Resmini E, Gomez-Anson B et al (2010) Magnetic resonance imaging as a predictor of response to somatostatin analogs in acromegaly after surgical failure. J Clin Endocrinol Metab 95:4973–4978. CrossRefPubMedGoogle Scholar
  5. 5.
    Heck A, Ringstad G, Fougner SL et al (2012) Intensity of pituitary adenoma on T2-weighted magnetic resonance imaging predicts the response to octreotide treatment in newly diagnosed acromegaly. Clin Endocrinol (Oxf) 77:72–78.
  6. 6.
    Shen M, Zhang Q, Liu W et al (2016) Predictive value of T2 relative signal intensity for response to somatostatin analogs in newly diagnosed acromegaly. Neuroradiology 58:1057–1065. CrossRefPubMedGoogle Scholar
  7. 7.
    Potorac I, Petrossians P, Daly AF et al (2016) T2-weighted MRI signal predicts hormone and tumor responses to somatostatin analogs in acromegaly. Endocr Relat Cancer 23:871–881. CrossRefPubMedGoogle Scholar
  8. 8.
    Lubner MG, Smith AD, Sandrasegaran K, Sahani DV, Pickhardt PJ (2017) CT texture analysis: definitions, applications, biologic correlates, and challenges. Radiographics 37:1483–1503.
  9. 9.
    Heck A, Emblem KE, Casar-Borota O, Bollerslev J, Ringstad G (2016) Quantitative analyses of T2-weighted MRI as a potential marker for response to somatostatin analogs in newly diagnosed acromegaly. Endocrine 52:333–343.
  10. 10.
    Tustison NJ, Avants BB, Cook PA et al (2010) N4ITK: improved N3 bias correction. IEEE Trans Med Imaging 29:1310–1320. CrossRefPubMedPubMedCentralGoogle Scholar
  11. 11.
    Collewet G, Strzelecki M, Mariette F (2004) Influence of MRI acquisition protocols and image intensity normalization methods on texture classification. Magn Reson Imaging 22:81–91. CrossRefPubMedGoogle Scholar
  12. 12.
    Shafiq-Ul-Hassan M, Zhang GG, Latifi K et al (2017) Intrinsic dependencies of CT radiomic features on voxel size and number of gray levels. Med Phys 44:1050–1062.
  13. 13.
    van Griethuysen JJM, Fedorov A, Parmar C et al (2017) Computational radiomics system to decode the radiographic phenotype. Cancer Res 77:e104–e107. CrossRefPubMedPubMedCentralGoogle Scholar
  14. 14.
    Kohavi R, John GH (1997) Wrappers for feature subset selection. Artif Intell 97:273–324. CrossRefGoogle Scholar
  15. 15.
    Bermejo P, Gamez JA, Puerta JM (2011) Improving incremental wrapper-based subset selection via replacement and early stopping. Intern J Pattern Recognit Artif Intell 25:605–625. CrossRefGoogle Scholar
  16. 16.
    Mwangi B, Tian TS, Soares JC (2014) A review of feature reduction techniques in neuroimaging. Neuroinformatics 12:229–244. CrossRefPubMedPubMedCentralGoogle Scholar
  17. 17.
    Cawley GC, Talbot NLC (2010) On over-fitting in model selection and subsequent selection bias in performance evaluation. J Mach Learn Res 11:2079–2107Google Scholar
  18. 18.
    Obari A, Sano T, Ohyama K et al (2008) Clinicopathological features of growth hormone-producing pituitary adenomas: difference among various types defined by cytokeratin distribution pattern including a transitional form. Endocr Pathol 19:82–91. CrossRefPubMedGoogle Scholar
  19. 19.
    Aha DW, Kibler D, Albert MK (1991) Instance-based learning algorithms. Mach Learn 6:37–66. CrossRefGoogle Scholar
  20. 20.
    Salzberg SL (1994) C4.5: Programs for Machine Learning by J. Ross Quinlan. Morgan Kaufmann Publishers, Inc., 1993. Mach Learn 16:235–240. CrossRefGoogle Scholar
  21. 21.
    Bradley AP (1997) The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recognit 30:1145–1159. CrossRefGoogle Scholar
  22. 22.
    Demšar J (2006) Statistical comparisons of classifiers over multiple data sets. J Mach Learn Res 7:1–30Google Scholar
  23. 23.
    Carlsen SM, Lund-Johansen M, Schreiner T et al (2008) Preoperative octreotide treatment in newly diagnosed acromegalic patients with macroadenomas increases cure short-term postoperative rates: a prospective, randomized trial. J Clin Endocrinol Metab 93:2984–2990. CrossRefPubMedGoogle Scholar
  24. 24.
    Mao ZG, Zhu YH, Tang HL et al (2010) Preoperative lanreotide treatment in acromegalic patients with macroadenomas increases short-term postoperative cure rates: a prospective, randomised trial. Eur J Endocrinol 162:661–666. CrossRefPubMedGoogle Scholar
  25. 25.
    Shen M, Shou X, Wang Y et al (2010) Effect of presurgical long-acting octreotide treatment in acromegaly patients with invasive pituitary macroadenomas: a prospective randomized study. Endocr J 57:1035–1044. CrossRefPubMedGoogle Scholar
  26. 26.
    Bacigaluppi S, Gatto F, Anania P et al (2016) Impact of pre-treatment with somatostatin analogs on surgical management of acromegalic patients referred to a single center. Endocrine 51:524–533. CrossRefPubMedGoogle Scholar
  27. 27.
    Kuhn M, Johnson K (2013) Over-fitting and model tuning. In: Applied predictive modeling. Springer New York, New York, pp 61–92CrossRefGoogle Scholar
  28. 28.
    Varma S, Simon R (2006) Bias in error estimation when using cross-validation for model selection. BMC Bioinformatics 7(91).

Copyright information

© European Society of Radiology 2018

Authors and Affiliations

  1. 1.Department of RadiologyIstanbul Training and Research HospitalIstanbulTurkey
  2. 2.Department of Radiology, Cerrahpasa Medical FacultyIstanbul University-CerrahpasaIstanbulTurkey
  3. 3.Department of Endocrinology and Metabolism, Cerrahpasa Medical FacultyIstanbul University-CerrahpasaIstanbulTurkey
  4. 4.Department of Pathology, Cerrahpasa Medical FacultyIstanbul University-CerrahpasaIstanbulTurkey
  5. 5.Department of Neurosurgery, Cerrahpasa Medical FacultyIstanbul University-CerrahpasaIstanbulTurkey

Personalised recommendations