Advertisement

Variable Ranking with PCA: Finding Multiparametric MR Imaging Markers for Prostate Cancer Diagnosis and Grading

  • Shoshana Ginsburg
  • Pallavi Tiwari
  • John Kurhanewicz
  • Anant Madabhushi
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6963)

Abstract

Although multiparametric (MP) MRI (MP–MRI) is a valuable tool for prostate cancer (CaP) diagnosis, considerable challenges remain in the ability to quantitatively combine different MRI parameters to train integrated, fused meta–classifiers for in vivo disease detection and characterization. To deal with the large number of MRI parameters, dimensionality reduction schemes such as principal component analysis (PCA) are needed to embed the data into a reduced subspace to facilitate classifier building. However, while features in the embedding space do not provide physical interpretability, direct feature selection in the high–dimensional space is encumbered by the curse of dimensionality. The goal of this work is to identify the most discriminating MP–MRI features for CaP diagnosis and grading based on their contributions in the reduced embedding obtained by performing PCA on the full MP–MRI feature space. In this work we demonstrate that a scheme called variable importance projection (VIP) can be employed in conjunction with PCA to identify the most discriminatory attributes. We apply our new PCA–VIP scheme to discover MP–MRI markers for discrimination between (a) CaP and benign tissue using 12 studies comprised of T2–w, DWI, and DCE MRI protocols and (b) high and low grade CaP using 36 MRS studies. The PCA–VIP score identified ADC values obtained from Diffusion and Gabor gradient texture features extracted from T2–w MRI as being most significant for CaP diagnosis. Our method also identified 3 metabolites that play a role in CaP detection—polyamine, citrate, and choline—and 4 metabolites that differentially express in low and high grade CaP: citrate, choline, polyamine, and creatine. The PCA–VIP scheme offers an alternative to traditional feature selection schemes that are encumbered by the curse of dimensionality.

Keywords

Principal Component Analysis Magnetic Resonance Spectroscopy Benign Tissue Variable Ranking Magnetic Resonance Spectroscopy Study 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    Jemal, A., et al.: Cancer Statistics. CA Cancer Journal for Clinicians 60(5), 277–300 (2010)CrossRefGoogle Scholar
  2. [2]
    Borboroglu, P.G., et al.: Extensive repeat transrectal ultrasound guided prostate biopsy in patients with previous benign sextant biopsies. The Journal of Urology 163(1), 158–162 (2000)CrossRefGoogle Scholar
  3. [3]
    Schiebler, M.L., et al.: Current role of MR imaging in the staging of adenocarcinoma of the prostate. Radiology 189(2), 339–352 (1993)CrossRefGoogle Scholar
  4. [4]
    Kurhanewicz, J., et al.: Combined magnetic resonance imaging and spectroscopic imaging appraoch to molecular imaging of prostate cancer. Journal of Magnetic Resonance Imaging 16, 451–463 (2002)CrossRefGoogle Scholar
  5. [5]
    Shukla–Dave, A., et al.: The utility of magnetic resonance imaging and spectroscopy for predicting insignificant prostate cancer: An initial analysis. BJU International 99, 786–793 (2007)CrossRefGoogle Scholar
  6. [6]
    Langer, D.L., et al.: Prostate tissue composition and MR measurements: Investigating the relationsips between ADC, T2, Ktrans, ve, and corresponding histological features. Radiology 255, 485–494 (2010)CrossRefGoogle Scholar
  7. [7]
    Klotz, L.: Active surveillance for prostate cancer: For whom? Journal of Clinical Oncology 23(32), 8165–8169 (2005)CrossRefGoogle Scholar
  8. [8]
    May, F., et al.: Limited value of endorectal magnetic resonance imaging and transrectal ultrasonography in the staging of clinically localized prostate cancer. BJU International 87(1), 66–69 (2001)CrossRefGoogle Scholar
  9. [9]
    Bonilla, J., et al.: Intra– and interobserver variability of MRI prostate volume measurements. Prostate 31(2), 98–102 (1997)CrossRefGoogle Scholar
  10. [10]
    Wetter, A., et al.: Combined MRI and MR spectroscopy of the prostate before radical prostatectomy. American Journal of Roentgenology 187, 724–730 (2006)CrossRefGoogle Scholar
  11. [11]
    Liu, X., et al.: Prostate cancer segmentation with simultaneous estimation of Markov random field parameters and class. IEEE Transactions on Medical Imaging 28(6), 906–915 (2009)CrossRefGoogle Scholar
  12. [12]
    Tiwari, P., et al.: A hierarchical spectral clustering and nonlinear dimensionality reduction scheme for detection of prostate cancer from magnetic resonance spectroscopy (MRS). Medical Physics 36(9), 3927–3939 (2009)CrossRefGoogle Scholar
  13. [13]
    Vos, P.C., et al.: Computer–assisted analysis of peripheral zone prostate lesions using T2–weighted and dynamic contrast enhanced T1–weighted MRI. Physics in Medicine and Biology 55(6), 1719 (2010)CrossRefGoogle Scholar
  14. [14]
    Viswanath, S., et al.: Enhanced multi–protocol analysis via intelligent supervised embedding (EMPrAvISE): Detecting prostate cancer on multi–parametric MRI. In: SPIE Medical Imaging, vol. 7963, pp.79630UGoogle Scholar
  15. [15]
    Tiwari, P., et al.: Multimodal wavelet embedding representation for data combination (MaWERiC): Integrating magnetic resonance imaging and spectroscopy for prostate cancer detection. Accepted to: NMR in MedicineGoogle Scholar
  16. [16]
    Duda, R., et al.: Pattern Classification, 2nd edn. Wiley-Interscience, Chichester (2000)Google Scholar
  17. [17]
    Kelm, B.M., et al.: Automated estimation of tumor probability in prostate magnetic resonance spectroscopic imaging: Pattern recognition vs quantification. Magnetic Resonance in Medicine 57, 150–159 (2007)CrossRefGoogle Scholar
  18. [18]
    Craig, A., et al.: Scaling and normalization effects in NMR spectroscopic metabonomic data sets. Analytical Chemistry 78, 2262–2267 (2006)CrossRefGoogle Scholar
  19. [19]
    Yan, H., et al.: Correntropy based feature selection using binary projection. Pattern Recognition 44(12), 2834–2843 (2011)CrossRefzbMATHGoogle Scholar
  20. [20]
    Chong, I.G., Jun, C.H.: Performance of some variable selection methods when multicollinearity is present. Chemometrics and Intelligent Laboratory Systems 78, 103–112 (2005)CrossRefGoogle Scholar
  21. [21]
    Esbensen, K.: Multivariate data analysis—in practice: An introduction to multivariate data analysis and experimental design. CAMO, Norway (2004)Google Scholar
  22. [22]
    Chappelow, J., et al.: Elastic registration of multimodal prostate MRI and histology via multiattribute combined mutual information. Medical Physics 38, 2005–2018 (2011)CrossRefGoogle Scholar
  23. [23]
    Madabhushi, A., et al.: Automated detection of prostatic adenocarcinoma from high resolution ex vivo MRI. IEEE Transactions on Medical Imaging 24(12), 1611–1625 (2005)CrossRefGoogle Scholar
  24. [24]
    Bovik, A.C., et al.: Multichannel texture analysis using localized spatial filters. IEEE Transactions on Pattern Analysis and Machine Intelligence 12(1), 55–73 (1990)CrossRefGoogle Scholar
  25. [25]
    Quinlan, J.R.: C4.5: Programs for Machine Learning. Morgan Kaufmann, San Francisco (1993)Google Scholar
  26. [26]
    Witten, I.H., Frank, E.: Data Mining: Practical Machine Learning Tools and Techniques. Elsevier, Amsterdam (2005)zbMATHGoogle Scholar
  27. [27]
    Nelson, S.J., Brown, T.: A new method for automatic quantification of 1–D spectra with low signal to noise ratio. Journal of Magnetic Resonance Imaging 84, 95–109 (1987)Google Scholar
  28. [28]
    Nelson, S.J.: Analysis of volume MRI and MR spectroscopic imaging data for the evaluation of patients with brain tumors. Magnetic Resonance in Medicine 46, 228–239 (2001)CrossRefGoogle Scholar
  29. [29]
    Devos, A., et al.: Classification of brain tumours using short echo time 1H MR spectra. Journal of Magnetic Resonance 170(1), 164–175 (2004)CrossRefGoogle Scholar
  30. [30]
    Swanson, M.G., et al.: 1H HR–MAS investigation of four potential markers for prostate cancer. Proc. Intl. Soc. Mag. Reson. Med. 9 (2001)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Shoshana Ginsburg
    • 1
  • Pallavi Tiwari
    • 1
  • John Kurhanewicz
    • 2
  • Anant Madabhushi
    • 1
  1. 1.Department of Biomedical EngineeringRutgers UniversityUSA
  2. 2.Department of RadiologyUniversity of CaliforniaSan FranciscoUSA

Personalised recommendations