Classification of Pancreatic Cysts in Computed Tomography Images Using a Random Forest and Convolutional Neural Network Ensemble

  • Konstantin DmitrievEmail author
  • Arie E. Kaufman
  • Ammar A. Javed
  • Ralph H. Hruban
  • Elliot K. Fishman
  • Anne Marie Lennon
  • Joel H. Saltz
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10435)


There are many different types of pancreatic cysts. These range from completely benign to malignant, and identifying the exact cyst type can be challenging in clinical practice. This work describes an automatic classification algorithm that classifies the four most common types of pancreatic cysts using computed tomography images. The proposed approach utilizes the general demographic information about a patient as well as the imaging appearance of the cyst. It is based on a Bayesian combination of the random forest classifier, which learns subclass-specific demographic, intensity, and shape features, and a new convolutional neural network that relies on the fine texture information. Quantitative assessment of the proposed method was performed using a 10-fold cross validation on 134 patients and reported a classification accuracy of 83.6%.



This research has been generously supported by The Marcus Foundation, Inc., and partially by NSF grants CNS0959979, IIP1069147, CNS1302246, NRT1633299, CNS1650499, IIS1527200, and NIH grant CA62924.


  1. 1.
    Cho, H.W., Choi, J.Y., Kim, M.J., Park, M.S., Lim, J.S., Chung, Y.E., Kim, K.W.: Pancreatic tumors: emphasis on CT findings and pathologic classification. Korean J. Radiol 12(6), 731–739 (2011)CrossRefGoogle Scholar
  2. 2.
    Criminisi, A., Shotton, J., Konukoglu, E.: Decision forests for classification, regression, density estimation, manifold learning and semi-supervised learning. Microsoft Research Cambridge, Technical report MSRTR-2011-114 5(6), 12 (2011)Google Scholar
  3. 3.
    Dmitriev, K., Gutenko, I., Nadeem, S., Kaufman, A.: Pancreas and cyst segmentation. In: Proceedings of SPIE Medical Imaging, p. 97842C (2016)Google Scholar
  4. 4.
    Ingalhalikar, M., Parker, W.A., Bloy, L., Roberts, T.P.L., Verma, R.: Using multiparametric data with missing features for learning patterns of pathology. In: Ayache, N., Delingette, H., Golland, P., Mori, K. (eds.) MICCAI 2012. LNCS, vol. 7512, pp. 468–475. Springer, Heidelberg (2012). doi: 10.1007/978-3-642-33454-2_58 CrossRefGoogle Scholar
  5. 5.
    Lennon, A.M., Wolfgang, C.L., Canto, M.I., Klein, A.P., Herman, J.M., Goggins, M., Fishman, E.K., Kamel, I., Weiss, M.J., Diaz, L.A., et al.: The early detection of pancreatic cancer: what will it take to diagnose and treat curable pancreatic neoplasia? Cancer Res. 74(13), 3381–3389 (2014)CrossRefGoogle Scholar
  6. 6.
    Maggioni, M., Katkovnik, V., Egiazarian, K., Foi, A.: Nonlocal transform-domain filter for volumetric data denoising and reconstruction. IEEE Trans. Image Process. 22(1), 119–133 (2013)MathSciNetCrossRefzbMATHGoogle Scholar
  7. 7.
    Raman, S.P., Chen, Y., Schroeder, J.L., Huang, P., Fishman, E.K.: CT texture analysis of renal masses: pilot study using random forest classification for prediction of pathology. Acad. Radiol. 21(12), 1587–1596 (2014)CrossRefGoogle Scholar
  8. 8.
    Raman, S.P., Schroeder, J.L., Huang, P., Chen, Y., Coquia, S.F., Kawamoto, S., Fishman, E.K.: Preliminary data using computed tomography texture analysis for the classification of hypervascular liver lesions: generation of a predictive model on the basis of quantitative spatial frequency measurements - a work in progress. J. Comput. Assist. Tomogr. 39(3), 383–395 (2015)Google Scholar
  9. 9.
    Sahani, D.V., Sainani, N.I., Blake, M.A., Crippa, S., Mino-Kenudson, M., del Castillo, C.F.: Prospective evaluation of reader performance on mdct in characterization of cystic pancreatic lesions and prediction of cyst biologic aggressiveness. AJR Am. J. Roentgenol. 197(1), W53–W61 (2011)CrossRefGoogle Scholar
  10. 10.
    Shin, H.C., Roth, H.R., Gao, M., Lu, L., Xu, Z., Nogues, I., Yao, J., Mollura, D., Summers, R.M.: Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning. IEEE Trans. Med. Imaging 35(5), 1285–1298 (2016)CrossRefGoogle Scholar
  11. 11.
    Yang, M., Kpalma, K., Ronsin, J.: A survey of shape feature extraction techniques. In: Yin, P.-Y. (ed.) Pattern Recognition Techniques, Technology and Applications. InTech (2008). doi: 10.5772/6237 Google Scholar
  12. 12.
    Zaheer, A., Pokharel, S.S., Wolfgang, C., Fishman, E.K., Horton, K.M.: Incidentally detected cystic lesions of the pancreas on CT: review of literature and management suggestions. Abdom. Imaging 38(2), 331–341 (2013)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  • Konstantin Dmitriev
    • 1
    Email author
  • Arie E. Kaufman
    • 1
  • Ammar A. Javed
    • 2
  • Ralph H. Hruban
    • 3
  • Elliot K. Fishman
    • 4
  • Anne Marie Lennon
    • 2
    • 5
  • Joel H. Saltz
    • 6
  1. 1.Department of Computer ScienceStony Brook UniversityStony BrookUSA
  2. 2.Department of SurgeryJohns Hopkins School of MedicineBaltimoreUSA
  3. 3.The Department of Pathology, The Sol Goldman Pancreatic Cancer Research CenterJohns Hopkins School of MedicineBaltimoreUSA
  4. 4.Department of RadiologyJohns Hopkins School of MedicineBaltimoreUSA
  5. 5.Division of Gastroenterology and HepatologyJohns Hopkins School of MedicineBaltimoreUSA
  6. 6.Department of Biomedical InformaticsStony Brook UniversityStony BrookUSA

Personalised recommendations