Advertisement

Ultrasound-Based Detection of Prostate Cancer Using Automatic Feature Selection with Deep Belief Networks

  • Shekoofeh Azizi
  • Farhad Imani
  • Bo Zhuang
  • Amir Tahmasebi
  • Jin Tae Kwak
  • Sheng Xu
  • Nishant Uniyal
  • Baris Turkbey
  • Peter Choyke
  • Peter Pinto
  • Bradford Wood
  • Mehdi Moradi
  • Parvin Mousavi
  • Purang Abolmaesumi
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9350)

Abstract

We propose an automatic feature selection framework for analyzing temporal ultrasound signals of prostate tissue. The framework consists of: 1) an unsupervised feature reduction step that uses Deep Belief Network (DBN) on spectral components of the temporal ultrasound data; 2) a supervised fine-tuning step that uses the histopathology of the tissue samples to further optimize the DBN; 3) a Support Vector Machine (SVM) classifier that uses the activation of the DBN as input and outputs a likelihood for the cancer. In leave-one-core-out cross-validation experiments using 35 biopsy cores, an area under the curve of 0.91 is obtained for cancer prediction. Subsequently, an independent group of 36 biopsy cores was used for validation of the model. The results show that the framework can predict 22 out of 23 benign, and all of cancerous cores correctly. We conclude that temporal analysis of ultrasound data can potentially complement multi-parametric Magnetic Resonance Imaging (mp-MRI) by improving the differentiation of benign and cancerous prostate tissue.

Keywords

Temporal ultrasound data deep learning deep belief network cancer diagnosis prostate cancer feature selection classification 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Bengio, Y., Lamblin, P., Popovici, D., et al.: Greedy layer-wise training of deep networks. Advances in Neural Information Processing Systems 19, 153 (2007)Google Scholar
  2. 2.
    Daoud, M.I., Mousavi, P., Imani, F., Rohling, R., Abolmaesumi, P.: Tissue classification using ultrasound-induced variations in acoustic backscattering features. IEEE Transactions on Biomedical Engineering 60(2), 310–320 (2013)CrossRefGoogle Scholar
  3. 3.
    Hinton, G.: A practical guide to training RBM. Momentum 9(1), 926 (2010)Google Scholar
  4. 4.
    Imani, F., et al.: Ultrasound-based characterization of prostate cancer: An in vivo clinical feasibility study. In: Mori, K., Sakuma, I., Sato, Y., Barillot, C., Navab, N. (eds.) MICCAI 2013, Part II. LNCS, vol. 8150, pp. 279–286. Springer, Heidelberg (2013)CrossRefGoogle Scholar
  5. 5.
    Khojaste, A., Imani, F., Moradi, M., Berman, D., et al.: Characterization of aggressive prostate cancer using ultrasound RF time series. SPIE Medical Imaging (2015)Google Scholar
  6. 6.
    Liao, S., Gao, Y., Oto, A., Shen, D.: Representation learning: A unified deep learning framework for automatic prostate MR segmentation. In: Mori, K., Sakuma, I., Sato, Y., Barillot, C., Navab, N. (eds.) MICCAI 2013, Part II. LNCS, vol. 8150, pp. 254–261. Springer, Heidelberg (2013)CrossRefGoogle Scholar
  7. 7.
    Margel, D., Yap, S.A., Lawrentschuk, N., Klotz, L., et al.: Impact of multiparametric endorectal coil prostate magnetic resonance imaging on disease reclassification among active surveillance candidates: a prospective cohort study. The Journal of Urology 187(4), 1247–1252 (2012)CrossRefGoogle Scholar
  8. 8.
    Moradi, M., Mousavi, P., Boag, A., Sauerbrei, E.E., Siemens, D., Abolmaesumi, P.: Augmenting detection of prostate cancer in transrectal ultrasound images using SVM and RF time series. IEEE Transactions on Biomedical Engineering 56(9), 2214–2224 (2009)CrossRefGoogle Scholar
  9. 9.
    de Rooij, M., Hamoen, E.H., Fütterer, J.J., Barentsz, J.O., Rovers, M.M.: Accuracy of multiparametric MRI for prostate cancer detection: a meta-analysis. American Journal of Roentgenology 202(2), 343–351 (2014)CrossRefGoogle Scholar
  10. 10.
    Tanaka, M., Okutomi, M.: A novel inference of a restricted boltzmann machine. In: International Conference on Pattern Recognition, pp. 1526–1531. IEEE (2014)Google Scholar
  11. 11.
    van Tulder, G., de Bruijne, M.: Learning features for tissue classification with the classification restricted boltzmann machine. In: Menze, B., Langs, G., Montillo, A., Kelm, M., Müller, H., Zhang, S., Cai, W(T.), Metaxas, D. (eds.) MCV 2014. LNCS, vol. 8848, pp. 47–58. Springer, Heidelberg (2014)Google Scholar
  12. 12.
    Uniyal, N., et al.: Ultrasound-based predication of prostate cancer in MRI-guided biopsy. In: Linguraru, M.G., Laura, C.O., Shekhar, R., Wesarg, S., Ballester, M.Á.G., Drechsler, K., Sato, Y., Erdt, M. (eds.) CLIP 2014. LNCS, vol. 8680, pp. 142–150. Springer, Heidelberg (2017)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  • Shekoofeh Azizi
    • 1
  • Farhad Imani
    • 1
  • Bo Zhuang
    • 1
  • Amir Tahmasebi
    • 2
  • Jin Tae Kwak
    • 3
  • Sheng Xu
    • 3
  • Nishant Uniyal
    • 1
  • Baris Turkbey
    • 4
  • Peter Choyke
    • 4
  • Peter Pinto
    • 4
  • Bradford Wood
    • 4
  • Mehdi Moradi
    • 5
  • Parvin Mousavi
    • 6
  • Purang Abolmaesumi
    • 1
  1. 1.The University of British ColumbiaVancouverCanada
  2. 2.Philips Research North AmericaBriarcliff ManorUSA
  3. 3.National Institutes of HealthBethesdaUSA
  4. 4.National Cancer InstituteBethesdaUSA
  5. 5.IBM Almaden Research CenterSan JoseUSA
  6. 6.Queen’s UniversityKingstonCanada

Personalised recommendations