Recovering Partially Sampled EEG Signals Using Learned Dictionaries

Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 390)

Abstract

Prior studies have proposedmethods to recover multi-channel EEG signal ensembles from their partially sampled entries. These techniques are only limited to multi-channel scenarios, especially when the number of channels is very large. Most biomedical signals (apart from EEG) are acquired from a single channel or from very few channels. Existing techniques cannot be employed to recover such signals from their partial samples. In this work, we propose a dictionary learning-based technique to overcome the problem. A sparsifying dictionary is learnt from the training examples; the trained dictionary is then used as a sparsifying transform in compressed sensing settings to recover the partially sampled test signal. Our proposed dictionary learning-based technique shows significant improvement in recovery over fixed sparsifying basis.

Keywords

Dictionary learning Compressed sensing Wireless body area network 

References

  1. 1.
    Aviyente, S.: Compressed sensing framework for EEG compression. In: IEEE Workshop on Statistical Signal Processing, pp. 181–184 (2007)Google Scholar
  2. 2.
    Abdulghani, A.M., Casson, A.J., Rodriguez-Villegas, E.: Quantifying the performance of compressive sensing on scalp EEG signals. In: International Symposium on Applied Sciences in Biomedical and Communication Technologies, pp. 1–5; 7–10 (2010)Google Scholar
  3. 3.
    Kamal, M.H., Shoaran, M., Leblebici, Y., Schmid, A., Vandergheynst, P.: Compressive multichannel cortical signal recording. In: Proceedings of the 38th International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Vancouver, Canada (2013)Google Scholar
  4. 4.
    Zhang, Z., Tzyy-Ping, J., Makeig, S., Rao, B.D.: Compressed sensing of EEG for wireless telemonitoring with low energy consumption and inexpensive hardware. IEEE Trans. Biomed. Eng. (accepted)Google Scholar
  5. 5.
    Mohsina, M., Majumdar, A.: Gabor based analysis prior formulation for EEG signal reconstruction. Biomed. Signal Process. Control 8(6), 951955 (2013)CrossRefGoogle Scholar
  6. 6.
    Majumdar, A., Ward, R.K.: Non-convex row-sparse MMV analysis prior formulation for EEG signal reconstruction. Biomed. Signal Process. Control 13, 142147 (2014)CrossRefGoogle Scholar
  7. 7.
    Shukla, A., Majumdar, A.: Row-sparse blind compressed sensing for reconstructing multi-channel EEG signals. Biomed. Signal Process. Control (accepted)Google Scholar
  8. 8.
    Majumdar, A., Gogna, A., Ward, R.: Low-rank matrix recovery approach for energy efficient EEG acquisition for wireless body area network. Sens. Special Issue State-of-the-art Sens. Technol. Can. 14(9), 15729–15748 (2014)Google Scholar
  9. 9.
    Cands, E., Romberg, J.: Sparsity and incoherence in compressive sampling. Inverse Prob. 23(3), 969 (2007)CrossRefGoogle Scholar
  10. 10.
    Hamner, B., Chavarriaga, R., Jos del, R.M.: Learning dictionaries of spatial and temporal EEG primitives for brain-computer interfaces. In: Workshop on Structured Sparsity: Learning and Inference, ICML (2011)Google Scholar
  11. 11.
    Zhou, W., Yang, Y., Yu, Z.: Discriminative dictionary learning for EEG signal classification in Brain-computer interface. In: Proceedings of the 12th International Conference on Control Automation Robotics and Vision (ICARCV), pp. 1582–1585 (2012)Google Scholar
  12. 12.
    Aharon, Michal, Elad, Michael, Bruckstein, Alfred: K-SVD: an algorithm for designing overcomplete dictionaries for sparse representation. Inverse Prob. 54(11), 4311–4322 (2006)Google Scholar
  13. 13.
  14. 14.
    Arora, S., Ge, R., Moitra, A.: New algorithms for learning incoherent and overcomplete dictionaries. arXiv:1308.6273
  15. 15.
    Barchiesi, D., Plumbley, M.D.: Learning incoherent dictionaries for sparse approximation using iterative projections and rotations. IEEE Trans. Signal Process. 61(8), 2055–2065 (2013)CrossRefGoogle Scholar

Copyright information

© Springer India 2016

Authors and Affiliations

  1. 1.Indraprastha Institute of Information Technology DelhiNew DelhiIndia

Personalised recommendations