專注力(Attention)在日常生活中扮演著重要的角色，影響著學習、工作和生活表現。過去已有多篇研究利用感覺刺激探討專注力，然而，在過往使用腦電圖(Electroencephalography, EEG)訊號進行專注力的研究中，主要注重於單一感覺刺激下的專注力分析，較少探討是否有與感知無關之專注力特徵。因此本研究設計一套藉由視覺刺激的心算任務(Mental Arithmetic, MA)與聽覺刺激的新異刺激任務(Oddball Paradigm)誘發使用者專注力之實驗流程，並使用靜息狀態(Resting-state, RS)、心算任務與新異刺激任務的腦波資料進行分析探討，希望藉由機器學習演算法找出與感知無關之專注力相關腦波生物標記，在未來能應用於專注力評估與檢測等領域。
故本論文使用頻帶功率(Band Power, BP)、碎形維度(Fractal Dimension, FD)與相干性(Coherence, Coh)進行特徵抽取，並利用費雪最佳特徵演算法進行特徵篩選，同時減少使用的電極數量和資料擷取時間，進而使用線性鑑別分析(Linear Discriminant Analysis, LDA)分類器，藉此從不同任務的最佳特徵中找出最佳共同特徵。實驗結果顯示：相干性在β和γ頻帶下有較多特徵符合條件，可能為專注力相關的生物標記；Katz碎形維度(Katz Fractal Dimension, KFD)的P3電極有符合條件，且進一步做統計分析後，與篩選共同特徵的條件相吻合；high-beta BP下的O1，在符合條件之15個雙特徵之特徵組合中，共出現11次，對high-beta BP下的O1特徵值做統計分析，也與篩選共同特徵的條件相吻合，KFD的P3電極和high-beta BP下的O1電極也可能是更客觀的專注力腦波標記。因此，本論文所發現的專注力特徵，未來有潛力可以進一步開發成為與感知無關之專注力腦波評估或量化的腦機介面特徵，並可對未來專注力研究與應用有更準確且專一之鑑別。

Attention plays an important role in daily life, influencing learning, work, and overall performance. Previous research has focused on investigating attention using sensory stimuli. However, in past studies using electroencephalography (EEG) signals to assess attention, the emphasis has primarily been on analyzing attention under single sensory stimulation, with fewer studies exploring attention features unrelated to perception. Therefore, this study designs an experimental procedure that induces attention using a mental arithmetic (MA) task with visual stimuli and an oddball paradigm with auditory stimuli. The EEG signals during resting-state (RS), MA task, and oddball task are analyzed to identify attention-related brainwave biomarkers using machine learning algorithms. The aim is to discover attention-related EEG biomarkers that are independent of sensory perception and potentially applicable in areas such as attention assessment and detection.
In this study, features such as band power (BP), fractal dimension (FD), and magnitude-squared coherence (Coh) are extracted. Fisher's optimal feature selection algorithm is employed to reduce the number of electrodes used and data acquisition time. Linear discriminant analysis (LDA) classifier is utilized to identify the best common features among different tasks. The experimental results indicate that coherence exhibits more features that meet the criteria in the beta and gamma bands, suggesting their potential as attention-related biomarkers. Katz fractal dimension (KFD) at the P3 electrode meets the criteria, and further statistical analysis aligns with the criteria for feature selection. In the high-beta BP, the O1 electrode appears 11 times among the 15 selected dual-feature combinations that meet the criteria. Statistical analysis of the O1 electrode's features in high-beta BP also matches the criteria for feature selection. Therefore, the P3 electrode's KFD and the O1 electrode in high-beta BP may serve as more objective attention-related EEG markers. The discovered attention features in this study have the potential to be further developed as perception-independent EEG features for assessing or quantifying attention in brain-computer interface applications. They can contribute to more accurate and specific discrimination in future attention research and applications.

[1] J. W. Choi, J. Choi, and S. Jo, “Reference bank multi-feature extraction for EEG-based concentration discrimination,” 2023 11th International Winter Conference on Brain-Computer Interface (BCI), 2023.
[2] B. Hamadicharef et al., “Learning EEG-based spectral-spatial patterns for attention level measurement,” 2009 IEEE International Symposium on Circuits and Systems, 2009.
[3] E. Wronka, J. Kaiser, AM. Coenen, “The auditory P3 from passive and active three-stimulus oddball paradigm,” Acta Neurobiol Exp (Wars), 2008.
[4] K. Ersin, O. Gundogdu, S. N. Kaya, D. Aykiri, and M. B. Serbetcioglu, “Investigation of the effects of auditory and visual stimuli on attention,” Heliyon, vol. 7, no. 7, 2021.
[5] C. T. Wu, D. G. Dillon, H. C. Hsu, S. Huang, E. Barrick, and Y. H. Liu, “Depression detection using relative EEG power induced by emotionally positive images and a conformal kernel support vector machine,” Applied Sciences, vol. 8, iss. 8, 2018.
[6] Q. Wang, O. Sourina, and M. K. Nguyen, “Fractal dimension based neurofeedback in serious games,” The Visual Computer, vol. 27, no. 4, pp. 299–309, 2011.
[7] W. K. So, S. W. Wong, J. N. Mak, and R. H. Chan, “An evaluation of mental workload with frontal EEG,” PLOS ONE, vol. 12, no. 4, 2017.
[8] R. Mahajan, C. A. Majmudar, S. Khatun, B. I. Morshed, and G. M. Bidelman, “NeuroMonitor ambulatory EEG device: Comparative Analysis and its application for Cognitive Load Assessment,” 2014 IEEE Healthcare Innovation Conference (HIC), 2014.
[9] J. Polich and A. Kok, “Cognitive and biological determinants of p300: An integrative review,” Biological Psychology, vol. 41, no. 2, pp. 103–146, 1995.
[10] J. Polich, “Updating p300: An integrative theory of P3A and p3b,” Clinical Neurophysiology, vol. 118, no. 10, pp. 2128–2148, 2007.
[11] J. V. Hurtado-Rincón, F. Restrepo, J. I. Padilla, H. F. Torres, and G. Castellanos-Dominguez, “Functional connectivity analysis using the oddball auditory paradigm for attention tasks,” Brain Informatics, pp. 99–108, 2018.
[12] E. Donchin and M. G. Coles, “Is the P300 component a manifestation of context updating?,” Behavioral and Brain Sciences, vol. 11, no. 03, p. 357, 1988.
[13] M. Senderecka, A. Grabowska, K. Gerc, J. Szewczyk, and R. Chmylak, “Event-related potentials in children with attention deficit hyperactivity disorder: An investigation using an auditory oddball task,” International Journal of Psychophysiology, vol. 85, no. 1, pp. 106–115, 2012.
[14] A. KOK, “On the utility of P3 amplitude as a measure of processing capacity,” Psychophysiology, vol. 38, no. 3, pp. 557–577, 2001.
[15] M. Kawasaki, K. Kitajo, and Y. Yamaguchi, “Dynamic links between Theta executive functions and alpha storage buffers in auditory and visual working memory,” European Journal of Neuroscience, 2010.
[16] H. Siamaknejad, W. S. Liew, and C. K. Loo, “Fractal dimension methods to determine optimum EEG electrode placement for concentration estimation,” Neural Computing and Applications, vol. 31, no. 3, pp. 945–953, 2017.
[17] Z. Yin and J. Zhang, “Identification of temporal variations in mental workload using locally-linear-embedding-based EEG feature reduction and support-vector-machine-based clustering and classification techniques,” Computer Methods and Programs in Biomedicine, vol. 115, no. 3, pp. 119–134, 2014.
[18] G. Sammer et al., “Relationship between regional hemodynamic activity and simultaneously recorded EEG-theta associated with mental arithmetic-induced workload,” Human Brain Mapping, vol. 28, no. 8, pp. 793–803, 2007.
[19] S. Dimitriadis, Y. Sun, N. Laskaris, N. Thakor, and A. Bezerianos, “Revealing cross-frequency causal interactions during a mental arithmetic task through symbolic transfer entropy: A novel vector-quantization approach,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 24, no. 10, pp. 1017-1028, 2016.
[20] S. I. Dimitriadis et al., “Cognitive workload assessment based on the tensorial treatment of EEG estimates of cross-frequency phase interactions,” Annals of Biomedical Engineering, vol. 43, no. 4, pp. 977–989, 2014.
[21] Q. Wang and O. Sourina, “Real-time mental arithmetic task recognition from EEG signals,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 21, no. 2, pp. 225-232, 2013
[22] C. Zhao et al., “An EEG-based mind controlled virtual-human obstacle-avoidance platform in three dimensional virtual environment,” 2017 8th International IEEE/EMBS Conference on Neural Engineering (NER), 2017.
[23] C. Lin, M. Jung, Y. Wu, H. She, and T. Jung, “Neural correlates of mathematical problem solving,” International Journal of Neural Systems, vol. 25, no. 2, p. 1550004, 2015.
[24] A. S. Dizaji and H. Soltanian-Zadeh, “Change-point analysis of single-trial EEG of auditory and visual oddball tasks,” 2016 24th Iranian Conference on Electrical Engineering (ICEE), 2016.
[25] S.-C. Yeh et al., “A multiplayer online car racing virtual-reality game based on internet of brains,” Journal of Systems Architecture, vol. 89, pp. 30–40, 2018.
[26] J. Downar, A. P. Crawley, D. J. Mikulis, and K. D. Davis, “The effect of task relevance on the cortical response to changes in visual and auditory stimuli: An event-related fmri study,” NeuroImage, vol. 14, no. 6, pp. 1256–1267, 2001.
[27] G.H. Klem and J.W. Lüders and H.H. Jasper and C. Elger. “The ten-twenty electrode system of the International Federation. The International Federation of Clinical Neurophysiology.” Electroencephalography and Clinical Neurophysiology, vol. 52, pp. 3–6, 1999.
[28] V. J. Lawhern et al., “EEGNet: A compact convolutional neural network for EEG-based brain–computer interfaces,” Journal of Neural Engineering, vol. 15, no. 5, p. 056013, 2018.
[29] T. Mullen et al., “Real-time modeling and 3D visualization of source dynamics and connectivity using wearable EEG,” 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2013.
[30] S. Mika, G. Ratsch, J. Weston, B. Scholkopf, and K. R. Mullers, “Fisher discriminant analysis with kernels,” Neural Networks for Signal Processing IX: Proceedings of the 1999 IEEE Signal Processing Society Workshop (Cat. No.98TH8468).
[31] A. Delorme and S. Makeig, “EEGLAB: An open source toolbox for analysis of single-trial EEG Dynamics including independent component analysis,” Journal of Neuroscience Methods, vol. 134, no. 1, pp. 9–21, 2004.