Unsupervised Clustering of Hyperspectral Images of Brain Tissues by Hierarchical Non-negative Matrix Factorization

Bangalore Ravi Kiran, Bogdan Stanciulescu, Jesus Angulo

2016

Abstract

Hyperspectral images of high spatial and spectral resolutions are employed to perform the challenging task of brain tissue characterization and subsequent segmentation for visualization of in-vivo images. Each pixel is high-dimensional spectrum. Working on the hypothesis of pure-pixels on account of high spectral resolution, we perform unsupervised clustering by hierarchical non-negative matrix factorization to identify the pure-pixel spectral signatures of blood, brain tissues, tumor and other materials. This subspace clustering was further used to train a random forest for subsequent classification of test set images constituent of in-vivo and ex-vivo images. Unsupervised hierarchical clustering helps visualize tissue structure in in-vivo test images and provides a inter-operative tool for surgeons. Furthermore the study also provide a preliminary study of the classification and sources of errors in the classification process.

References

  1. Bruna, J. and Mallat, S. (2013). Invariant scattering convolution networks. IEEE Trans. Pattern Anal. Mach. Intell., 35(8):1872-1886.
  2. Dvorak, H. F. (2003). How tumors make bad blood vessels and stroma. The American journal of pathology, 162(6):1747-1757.
  3. Gillis, N., Kuang, D., and Park, H. (2015). Hierarchical clustering of hyperspectral images using rank-two nonnegative matrix factorization. IEEE T. Geoscience and Remote Sensing, 53(4):2066-2078.
  4. Gillis, N. and Luce, R. (2014). Robust near-separable nonnegative matrix factorization using linear optimization. Journal of Machine Learning Research, 15:1249-1280.
  5. Kuang, D. (2014). Nonnegative matrix factorization for clustering. PhD thesis, Georgia Institute of Technology.
  6. Kukreti, S., Cerussi, A. E., Tanamai, W., Hsiang, D., Tromberg, B. J., and Gratton, E. (2010). Characterization of metabolic differences between benign and malignant tumors: High-spectral-resolution diffuse optical spectroscopy. Radiology, 254(1):277-284.
  7. Liu, Z., Wang, H., and Li, Q. (2011). Tongue tumor detection in medical hyperspectral images. Sensors, 12(1):162-174.
  8. Lu, G. and Fei, B. (2014). Medical hyperspectral imaging: a review. Journal of biomedical optics, 19(1):010901- 010901.
  9. Lu, G., Halig, L., Wang, D., Qin, X., Chen, Z. G., and Fei, B. (2014). Spectral-spatial classification for noninvasive cancer detection using hyperspectral imaging. Journal of biomedical optics, 19(10):106004-106004.
  10. Ma, W.-K., Bioucas-Dias, J., Chan, T.-H., Gillis, N., Gader, P., Plaza, A., Ambikapathi, A., and Chi, C.-Y. (2014). A signal processing perspective on hyperspectral unmixing: Insights from remote sensing. Signal Processing Magazine, IEEE, 31(1):67-81.
  11. Panasyuk, S., Yang, S., Faller, D., Ngo, D., Lew, R., Freeman, J., and Rogers, A. (2007). Medical hyperspectral imaging to facilitate residual tumor identification during surgery. Cancer Biol Ther, pages 439-46.
  12. Peswani, D. L. (2007). Detection of Positive Cancer Margins Intra-operatively During Nephrectomy and Prostatectomy Using Optical Reflectance Spectroscopy. ProQuest.
  13. Sahu, A., McGoverin, C., Pleshko, N., Sorenmo, K., and Won, C.-H. (2013). Hyperspectral imaging system to discern malignant and benign canine mammary tumors. In SPIE Defense, Security, and Sensing, pages 87190W-87190W. International Society for Optics and Photonics.
  14. Schulder, M. and Carmel, P. W. (2003). Intraoperative magnetic resonance imaging: impact on brain tumor surgery. Cancer Control, 10(2):115-124.
  15. Tibshirani, R. and Walther, G. (2005). Cluster validation by prediction strength. Journal of Computational and Graphical Statistics, 14(3):511-528.
  16. Utzinger, U., Brewer, M., Silva, E., Gershenson, D., Blast, R. C., Follen, M., and Richards-Kortum, R. (2001). Reflectance spectroscopy for in vivo characterization of ovarian tissue. Lasers in surgery and medicine, 28(1):56-66.
Download


Paper Citation


in Harvard Style

Kiran B., Stanciulescu B. and Angulo J. (2016). Unsupervised Clustering of Hyperspectral Images of Brain Tissues by Hierarchical Non-negative Matrix Factorization . In Proceedings of the 9th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 2: BIOIMAGING, (BIOSTEC 2016) ISBN 978-989-758-170-0, pages 77-84. DOI: 10.5220/0005697600770084


in Bibtex Style

@conference{bioimaging16,
author={Bangalore Ravi Kiran and Bogdan Stanciulescu and Jesus Angulo},
title={Unsupervised Clustering of Hyperspectral Images of Brain Tissues by Hierarchical Non-negative Matrix Factorization},
booktitle={Proceedings of the 9th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 2: BIOIMAGING, (BIOSTEC 2016)},
year={2016},
pages={77-84},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0005697600770084},
isbn={978-989-758-170-0},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 9th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 2: BIOIMAGING, (BIOSTEC 2016)
TI - Unsupervised Clustering of Hyperspectral Images of Brain Tissues by Hierarchical Non-negative Matrix Factorization
SN - 978-989-758-170-0
AU - Kiran B.
AU - Stanciulescu B.
AU - Angulo J.
PY - 2016
SP - 77
EP - 84
DO - 10.5220/0005697600770084