Determining the Optimal Number of MEG Trials: A Machine Learning and Speech Decoding Perspective
Advancing the knowledge about neural speech mechanisms is critical for developing next-generation, faster brain computer interface to assist in speech communication for the patients with severe neurological conditions (e.g., locked-in syndrome). Among current neuroimaging techniques, Magnetoencephalography (MEG) provides direct representation for the large-scale neural dynamics of underlying cognitive processes based on its optimal spatiotemporal resolution. However, the MEG measured neural signals are smaller in magnitude compared to the background noise and hence, MEG usually suffers from a low signal-to-noise ratio (SNR) at the single-trial level. To overcome this limitation, it is common to record many trials of the same event-task and use the time-locked average signal for analysis, which can be very time consuming. In this study, we investigated the effect of the number of MEG recording trials required for speech decoding using a machine learning algorithm. We used a wavelet filter for generating the denoised neural features to train an Artificial Neural Network (ANN) for speech decoding. We found that wavelet based denoising increased the SNR of the neural signal prior to analysis and facilitated accurate speech decoding performance using as few as 40 single-trials. This study may open up the possibility of limiting MEG trials for other task evoked studies as well.
KeywordsMEG Speech Wavelets Artificial Neural Network
This project was supported by the University of Texas System Brain Research grant and by the National Institutes of Health (NIH) under award number R03 DC013990. We thank Dr. Mark McManis, Dr. Ted Mau, Dr. Angel W. Hernandez-Mulero, Kanishk Goel and the volunteering participants.
- 5.Duffy, J.: Motor Speech Disorders Substrates, Differential Diagnosis, and Management, 3rd edn, p. 295. Elsevier, St. Louis (2012)Google Scholar
- 7.Wolpaw, J.R., Mcfarland, D.: Control of a two-dimensional movement signal by a noninvasive brain-computer interface in humans. PNAS 51, 49–54 (2004)Google Scholar
- 11.Leuthardt, E.C., Cunningham, J., Barbour, D.: Brain-computer interface research. In: Guger, C., Allison, B., Edlinger, G. (eds.) Towards a Speech BCI Using ECoG. SpringerBriefs in Electrical and Computer Engineering, pp. 93–110. Springer, Heidelberg (2013). https://doi.org/10.1007/978-3-642-36083-1_10CrossRefGoogle Scholar
- 12.Herff, C., et al.: Brain-to-text: decoding spoken phrases from phone representations in the brain. Front. Neurosci. 9(217), 1–11 (2005)Google Scholar
- 13.Dash, D., Abrol, B., Sao, A., Biswal, B.: The model order limit: deep sparse factorization for resting brain. In: IEEE 15th International Symposium on Biomedical Imaging (ISBI), pp. 1244–1247 (2018)Google Scholar
- 14.Dash, D., Abrol, B., Sao, A., Biswal, B.: Spatial sparsification and low rank projection for fast analysis of multi-subject resting state fMRI data. In: IEEE 15th International Symposium on Biomedical Imaging (ISBI), pp. 1280–1283 (2018)Google Scholar
- 18.Wang, J., Kim, M., Hernandez-Mulero, A.H., Heitzman, D., Ferrari, P.: Towards decoding speech production from single-trial Magnetoencephalography (MEG) signals. In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 3036–3040 (2017)Google Scholar
- 19.Dash, D., Ferrari, P., Malik, S., Wang, J.: Overt speech retrieval from neuromagnetic signals using wavelets and artificial neural networks. In: IEEE Global Conference on Signal and Information Processing (GlobalSIP) (2018)Google Scholar
- 21.Attal, Y., et al.: Modelling and detecting deep brain activity with MEG and EEG. IRBM - Biomed. Eng. Res. 30, 133 (2009)Google Scholar
- 25.Dash. D., Kim, M., Ferrari, P., Wang, J.: Brain activation pattern analysis for speech production decoding from MEG signals. In: 25th Annual meeting of Biomedical Engineering Society (BMES) (2018)Google Scholar
- 26.Tadel, F., Baillet, S., Mosher, J.C., Pantazis, D., Leahy, R.M.: Brainstorm: a user-friendly Application for MEG/EEG analysis. Comput. Intell. Neurosci. 8, 8:1–8:13 (2011)Google Scholar