Generalização do aumento de dados para reduzir o número de épocas para calcular a média
DOI:
https://doi.org/10.5935/jetia.v10i48.1201Resumo
A forma de onda derivada da média de vários registros de sinais de EEG durante a estimulação representa um Potencial Relacionado a Eventos (ERP). Quando são empregados estímulos sensoriais, os potenciais resultantes são denominados potenciais evocados (EPs). Os PEs encontram aplicações em diversos domínios de pesquisa e ambientes clínicos, servindo como uma ferramenta valiosa na neurociência e na medicina devido à sua versatilidade em oferecer percepções objetivas sobre a função cerebral. No entanto, o método convencional de média de sinal usado para extrair EPs tem limitações inerentes, como a necessidade de vários testes para garantir a confiabilidade e maximizar a relação sinal-ruído (SNR). Isso exige mais tempo para o registro e o processamento dos dados. Além disso, a confiabilidade das respostas registradas pode ser comprometida devido à habituação do sujeito ao estímulo. Para enfrentar esses desafios, este estudo tem como objetivo aprimorar a SNR na extração de PE empregando o aumento de dados, reduzindo assim o número de registros necessários para o cálculo da média. O método proposto demonstra uma melhoria notável de aproximadamente 9,77 ± 2,65 dB em comparação com a média de sinal tradicional com o mesmo número de registros. Este estudo conclui que o aumento criterioso dos dados permite estimativas aprimoradas de SNR sem a necessidade de novos registros extensos.
Downloads
Referências
P. E. Clayson, C. J. Brush, y G. Hajcak, “Data quality and reliability metrics for event-related potentials (ERPs): The utility of subject-level reliability”, Int. J. Psychophysiol., vol. 165, pp. 121–136, jul. 2021, doi: 10.1016/j.ijpsycho.2021.04.004.
J. A. Carballo-Gonzalez, P. Valdes-Sosa, y M. Valdes-Sosa, “Detection of Event Related Potentials”, Int. J. Neurosci., vol. 46, núm. 3–4, pp. 109–122, ene. 1989, doi: 10.3109/00207458908986247.
A. Taboada-Crispi et al., “Simulation of Event-Related Potentials in a wide sense for assessment of detection∕estimation algorithms”, en AIP Conference Proceedings, 2008, vol. 1060, pp. 7–10. doi: 10.1063/1.3037119.
F. Congedo, Marco and Lopes da Silva, “Event-related potentials: General aspects of methodology and quantification”, en Niedermeyer’s Electroencephalography: Basic Principles, Clinical Applications, and Related Fields, D. L. S. and F. H. L. da Silva, Ed. Oxford University Press, 2018. [En línea]. Disponible en: https://hal.archives-ouvertes.fr/hal-01953600.
E. Laciar y R. Jane, “An improved weighted signal averaging method for high-resolution ECG signals”, en Computers in Cardiology 2001. Vol.28 (Cat. No.01CH37287), 2001, pp. 69–72. doi: 10.1109/CIC.2001.977593.
T. Pander, S. Pietraszek, y T. Przybyła, Information Technologies in Biomedicine, vol. 284, núm. 4. Cham: Springer International Publishing, 2014. doi: 10.1007/978-3-319-06596-0.
I. Torres-Rodríguez, C. A. Ferrer, E. Velarde-Reyes, y A. Taboada-Crispi, “Improving ensemble averaging by epoch detrending in evoked potentials”, en Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2019, vol. 11401 LNCS, pp. 707–714. doi: 10.1007/978-3-030-13469-3_82.
J. M. Leski, “Robust weighted averaging [of biomedical signals]”, IEEE Trans. Biomed. Eng., vol. 49, núm. 8, pp. 796–804, ago. 2002, doi: 10.1109/TBME.2002.800757.
C. E. Davila y M. S. Mobin, “Weighted averaging of evoked potentials”, IEEE Trans. Biomed. Eng., vol. 39, núm. 4, pp. 338–345, abr. 1992, doi: 10.1109/10.126606.
Z. Leonowicz, J. Karvanen, y S. L. Shishkin, “Trimmed estimators for robust averaging of event-related potentials”, J. Neurosci. Methods, vol. 142, núm. 1, pp. 17–26, mar. 2005, doi: 10.1016/j.jneumeth.2004.07.008.
K. Kotowski, K. Stapor, y J. Leski, “Improved robust weighted averaging for event-related potentials in EEG”, Biocybern. Biomed. Eng., vol. 39, núm. 4, pp. 1036–1046, oct. 2019, doi: 10.1016/j.bbe.2019.09.002.
D. X. Wang y C. E. Davila, “Subspace Averaging of Auditory Evoked Potentials”, en 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), jul. 2019, pp. 656–659. doi: 10.1109/EMBC.2019.8857818.
P. Prado-Gutierrez, E. Martínez-Montes, A. Weinstein, y M. Zañartu, “Estimation of auditory steady-state responses based on the averaging of independent EEG epochs”, PLoS One, vol. 14, núm. 1, p. e0206018, ene. 2019, doi: 10.1371/journal.pone.0206018.
S. M. K. K. Madsen, J. M. Harte, C. Elberling, y T. Dau, “Accuracy of averaged auditory brainstem response amplitude and latency estimates”, Int. J. Audiol., vol. 57, núm. 5, pp. 345–353, may 2018, doi: 10.1080/14992027.2017.1381770.
C. Elberling y M. Don, “Quality Estimation of Averaged Auditory Brainstem Responses”, Scand. Audiol., vol. 13, núm. 3, pp. 187–197, ene. 1984, doi: 10.3109/01050398409043059.
H. Litvan et al., “Comparison of Conventional Averaged and Rapid Averaged, Autoregressive-based Extracted Auditory Evoked Potentials for Monitoring the Hypnotic Level during Propofol Induction”, Anesthesiology, vol. 97, núm. 2, pp. 351–358, ago. 2002, doi: 10.1097/00000542-200208000-00011.
D. L. Jewett y J. S. Williston, “Auditory-evoked far fields averaged from the scalp of humans”, Brain, vol. 94, núm. 4, pp. 681–696, 1971, doi: 10.1093/brain/94.4.681.
N. Acır, Y. Erkan, y Y. A. Bahtiyar, “Auditory brainstem response classification for threshold detection using estimated evoked potential data: comparison with ensemble averaged data”, Neural Comput. Appl., vol. 22, núm. 5, pp. 859–867, abr. 2013, doi: 10.1007/s00521-011-0776-2.
M. Don, C. Elberling, y M. Waring, “Objective Detection of Averaged Auditory Brainstem Responses”, Scand. Audiol., vol. 13, núm. 4, pp. 219–228, ene. 1984, doi: 10.3109/01050398409042130.
L. Sörnmo y P. Laguna, Bioelectrical Signal Processing in Cardiac and Neurological Applications, vol. 8. Elsevier, 2005. doi: 10.1016/B978-0-12-437552-9.X5000-4.
E. Perez-Abalo, Maria Cecilia and Rodriguez-Davila, Ernesto Luis and Sanchez-Castillo, Manuel and Sotero-Diaz, Roberto Carlos and Torres-Fortuny, Alejandro and Santos-Febles, “Method and apparatus for the objective detection of auditive disorders”, US Patent 9,345,419, 2016 [En línea]. Disponible en: https://patents.google.com/patent/US9345419B2/en.
O. Rompelman y H. H. Ros, “Coherent averaging technique: A tutorial review Part 1: Noise reduction and the equivalent filter”, J. Biomed. Eng., vol. 8, núm. 1, pp. 24–29, ene. 1986, doi: 10.1016/0141-5425(86)90026-9.
M. A. Chesnaye, “Optimising Objective Detection Methods for the Auditory Brainstem Response”, University of Southampton, 2019. [En línea]. Disponible en: https://eprints.soton.ac.uk/428621/
K. Rennert et al., “British Columbia Early Hearing Program (BCEHP) Audiology Assessment Protocol”. 2012. doi: 10.13140/2.1.2977.5683.
I. Torres-Rodríguez, “Estimadores Robustos de Promediado para la Detección de Potenciales Relacionados con Eventos”, Universidad Central “Marta Abreu” de Las Villas, 2014.
I. Torres-Rodríguez, C. A. Ferrer-Riesgo, J. C. Oliva Pérez, y A. Taboada-Crispi, “Performance of Different Average Methods for the Automatic Detection of Evoked Potentials”, en Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11896 LNCS, 2019, pp. 629–636. doi: 10.1007/978-3-030-33904-3_59.
I. Torres-Rodríguez, C. A. Ferrer-Riesgo, M. M. P. de Morales Artiles, y A. Taboada-Crispi, “Performance Evaluation of Average Methods in the Time Domain Using Quality Measures for Automatic Detection of Evoked Potentials”, en IFMBE Proceedings, vol. 75, Springer, 2020, pp. 12–20. doi: 10.1007/978-3-030-30648-9_2.
C. A. Ferrer, I. Torres-Rodríguez, A. Taboada-Crispi, y E. Nöth, “Analytical Solution for the Optimal Addition of an Item to a Composite of Scores for Maximum Reliability”, en Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11896 LNCS, Springer, 2019, pp. 408–416. doi: 10.1007/978-3-030-33904-3_38.
K. Zhang et al., “Data Augmentation for Motor Imagery Signal Classification Based on a Hybrid Neural Network”, Sensors, vol. 20, núm. 16, p. 4485, ago. 2020, doi: 10.3390/s20164485.
C. He, J. Liu, Y. Zhu, y W. Du, “Data Augmentation for Deep Neural Networks Model in EEG Classification Task: A Review”, Frontiers in Human Neuroscience, vol. 15. Frontiers Media S.A., el 17 de diciembre de 2021. doi: 10.3389/fnhum.2021.765525.
R. Zhang et al., “ERP-WGAN: A data augmentation method for EEG single-trial detection”, J. Neurosci. Methods, vol. 376, p. 109621, jul. 2022, doi: 10.1016/j.jneumeth.2022.109621.
B. K. Iwana y S. Uchida, “An empirical survey of data augmentation for time series classification with neural networks”, PLoS One, vol. 16, núm. 7, p. e0254841, jul. 2021, doi: 10.1371/journal.pone.0254841.
M. M. Krell, A. Seeland, y S. K. Kim, “Data Augmentation for Brain-Computer Interfaces: Analysis on Event-Related Potentials Data”, arXiv Prepr. arXiv1801.02730, ene. 2018, doi: https://doi.org/10.48550/arXiv.1801.02730.
I. Torres-Rodríguez, R. Díaz-Amador, B. Peón-Pérez, A. Hurtado Armas, y A. Taboada-Crispi, “Realistic Simulation of Event-Related Potentials and Their Usual Noise and Interferences for Pattern Recognition”, en Pattern Recognition, 2023, pp. 201–210.
