Automatic detection of developmental coordination disorder from power spectral densities of electroencephalography (EEG) signals using deep learning model

Öz

Developmental coordination disorder is a neurodevelopmental disorder characterized by a marked deterioration in the development of motor and coordination skills that significantly impairs daily activity and academic performance. Early diagnosis is very important for medical intervention. Accurate diagnosis of the disease requires extensive testing and long-term observations. These tests and observations can be time consuming, expensive, incomplete, inaccurate and subjective. EEG signals are a method used to monitor brain activity used in early diagnosis. EEG is widely used in the diagnosis of diseases due to its advantages such as being non-invasive, being based on findings, being less costly and getting results in a short time. In this study, an EEG-based deep learning model is presented to support experts in detecting developmental coordination disorder in children from EEG signals. The dataset consists of EEG signals recorded from 16 children without developmental coordination disorder and 16 children with developmental coordination disorder. First of all, power spectral density values of frequencies between 1-49 hertz of EEG signals were calculated separately by using periodogram, welch and multitaper spectral analysis methods. For each of the three different spectral analysis methods, 49 feature vectors were extracted. Then, the performances of support vector machine (SVM), random forest (RF), k-nearest neighbor (kNN) and long-short-term memory (LSTM) algorithms are compared using extracted feature vectors. After the comparison, the model integrating the welch spectral analysis and the LSTM deep learning algorithm showed the highest performance as a result of the experiments. The proposed deep learning model achieved promising performance with 97.20% accuracy, 0.984 sensitivity, 0.959 specificity, 0.962 precision, 0.973 f1-score and 0.944 Matthews correlation coefficient (MCC). The study is a rare attempt in which a deep learning model is used in the effective diagnosis of automatic developmental coordination disorder by analyzing EEG signals and provides evidence of the superiority of deep learning algorithms over traditional machine learning algorithms.

Anahtar Kelimeler

• Deep learning
• developmental coordination disorder
• EEG signal processing
• spectral analysis
• long-short-term memory networks

Elektroensefalografi (EEG) sinyallerinin güç spektral yoğunlukları kullanılarak gelişimsel koordinasyon bozukluğunun derin öğrenme modeli ile otomatik tespiti

Öz

Gelişimsel koordinasyon bozukluğu, günlük aktivite ve akademik performansı önemli ölçüde engelleyen motor ve koordinasyon becerilerinin gelişiminde belirgin bir bozulma ile karakterize nöro-gelişimsel bir hastalıktır. Tıbbi müdahale için erken tanı çok önemlidir. Hastalığın doğru teşhisi kapsamlı testler ve uzun vadeli gözlemler gerektirir. Bu testler ve gözlemler zaman alıcı, pahalı, eksik, yanlış ve sübjektif olabilir. EEG sinyalleri, erken tanıda kullanılan beyin aktivitesini izlemek için kullanılan bir yöntemdir. EEG invaziv olmaması, bulgulara dayalı olması, daha az maliyetli olması ve kısa sürede sonuç alabilmesi gibi avantajları nedeniyle hastalıkların tanısında yaygın olarak kullanılmaktadır. Bu çalışmada EEG sinyallerinden çocuklarda gelişimsel koordinasyon bozukluğunun tespitinde uzmanlara destek olmak amacıyla EEG tabanlı bir derin öğrenme modeli sunulmaktadır. Veriseti 16 gelişimsel koordinasyon bozukluğu olmayan ve 16 gelişimsel koordinasyon bozukluğu olan çocuktan kaydedilen EEG sinyallerinden oluşmaktadır. Öncelikle periodogram, welch ve multitaper spektral analiz yöntemleri kullanılarak EEG sinyallerinin 1-49 hertz arasındaki frekanslarının güç spektral yoğunluk değerleri ayrı ayrı hesaplanmıştır. Üç farklı spektral analiz yöntemlerinin her biri için 49 özellik vektörü çıkarılmıştır. Daha sonra, çıkarılan öznitelik vektörleri kullanılarak destek vektör makinesi (SVM), rastgele orman (RF), k-en yakın komşu (kNN) ve uzun-kısa süreli bellek (LSTM) algoritmalarının performansları karşılaştırılmıştır. Karşılaştırma sonrasında welch spektral analizi ile LSTM derin öğrenme algoritmasını bütünleştiren model, deneyler sonucunda en yüksek performansı göstermiştir. Önerilen derin öğrenme modeli, %97,20 doğruluk, 0,984 duyarlılık, 0.959 özgüllük, 0,962 kesinlik, 0,973 f1-skoru ve 0,944 Matthews korelasyon katsayısı (MCC) değerleri ile umut verici bir performans elde etmiştir. Çalışma EEG sinyallerini analiz ederek otomatik gelişimsel koordinasyon bozukluğunun efektif tanısında derin öğrenme modelinin kullanıldığı nadir bir girişimdir ve geleneksel makine öğrenmesi algoritmalarına göre derin öğrenme algoritmalarının üstünlüğüne dair kanıt sağlamaktadır.

Anahtar Kelimeler

• Derin öğrenme
• gelişimsel koordinasyon bozukluğu
• EEG sinyal işleme
• spektral analiz
• uzun-kısa süreli bellek ağları

Kaynakça

1. American Psychiatric Association, Diagnostic and statistical manual of mental disorders (DSM, 5th edn.). Arlington, VA: American Psychiatric Association, 2013.
2. Lingam, R., Hunt, L., Golding, J., Jongmans, M., Emond, A., Prevalence of developmental coordination disorder using the DSM-IV at 7 years of age: A UK population–based study. Pediatrics, 123(4), e693-e700, 2009.
3. Kirby, A., Williams, N., Thomas, M., Hill, E. L., Self-reported mood, general health, wellbeing and employment status in adults with suspected DCD. Research in Developmental Disabilities, 34(4), 1357-1364. 2013.
4. Sumner, E., Hutton, S. B., Kuhn, G., Hill, E. L., Oculomotor atypicalities in developmental coordination disorder. Developmental Science, 21(1), e12501, 2018.
5. Draghi, T. T. G., Cavalcante Neto, J. L., Rohr, L. A., Jelsma, L. D., Tudella, E., Symptoms of anxiety and depression in children with developmental coordination disorder: a systematic review. Jornal de pediatria, 96, 08-19, 2020.
6. Izadi-Najafabadi, S., Ryan, N., Ghafooripoor, G., Gill, K., Zwicker, J. G., Participation of children with developmental coordination disorder. Research in developmental disabilities, 84, 75-84, 2019.
7. Brons, A., de Schipper, A., Mironcika, S., Toussaint, H., Schouten, B., Bakkes, S., Kröse, B., Assessing children’s fine motor skills with sensor-augmented toys: machine learning approach. Journal of Medical Internet Research, 23(4), e24237, 2021.
8. Baxter, P., Distinguishing ataxia from developmental coordination disorder. Developmental Medicine & Child Neurology, 62(1), 11-11, 2020.

9. Li, R., Fu, H., Zheng, Y., Lo, W. L., Yu, J. J., Sit, C. H., Chi, Z., Song, Z., Wen, D., Automated fine motor evaluation for developmental coordination disorder. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 27(5), 963-973, 2019.
10. Yıldırım, C., Acar, G., Polat, M. G., Mete, E., Kaygusuz, R., Neuroimaging in developmental coordination disorder. Turkish Journal of Neurology, 27(1), 5-16, 2021.
11. Brady, D., Leonard, H. C., Developmental coordination disorder. Chapter in the Oxford Handbook of Developmental Cognitive Neuroscience, 2019.
12. Aslan, Z., Migraine detection from EEG signals using tunable Q-factor wavelet transform and ensemble learning techniques. Physical and Engineering Sciences in Medicine, 44(4), 1201-1212, 2021.
13. Gomez-Pilar, J., García-Azorín, D., Gomez-Lopez-de-San-Roman, C., Guerrero, Á. L., Hornero, R., Exploring EEG spectral patterns in episodic and chronic migraine during the interictal state: determining frequencies of interest in the resting state. Pain Medicine, 21(12), 3530-3538, 2020.
14. Hyde, C., Fuelscher, I., Williams, J., Neurophysiological approaches to understanding motor control in DCD: current trends and future directions. Current Developmental Disorders Reports, 6(2), 78-86, 2019.
15. Fong, S. S., Ng, S. S., Chung, L. M., Ki, W. Y., Chow, L. P., Macfarlane, D. J., Direction-specific impairment of stability limits and falls in children with developmental coordination disorder: implications for rehabilitation. Gait & posture, 43, 60-64, 2016.
16. Tsai, C. L., Wang, C. H., Tseng, Y. T., Effects of exercise intervention on event-related potential and task performance indices of attention networks in children with developmental coordination disorder. Brain and cognition, 79(1), 12-22, 2012.
17. Martinez-Manzanera, O., Lawerman, T. F., Blok, H. J., Lunsing, R. J., Brandsma, R., Sival, D. A., Maurits, N. M., Instrumented finger-to-nose test classification in children with ataxia or developmental coordination disorder and controls. Clinical Biomechanics, 60, 51-59, 2018.
18. Buettner, R., Buechele, M., Grimmeisen, B., Ulrich, P., Machine learning based diagnostics of developmental coordination disorder using electroencephalographic data, Proceedings of the 54th Hawaii International Conference on System Sciences, 3426-3435, 2021.
19. Blank, R., Barnett, A. L., Cairney, J., Green, D., Kirby, A., Polatajko, H., Rosenblum, S., Smits-Engelsman, B., Sugden, D., Wilson, P., Vinçon, S. International clinical practice recommendations on the definition, diagnosis, assessment, intervention, and psychosocial aspects of developmental coordination disorder. Developmental Medicine & Child Neurology, 61(3), 242-285, 2019.
20. Vařeka, L., Brůha, P., Mouček, R., Mautner, P., Čepička, L., Holečková, I., Developmental coordination disorder in children–experimental work and data annotation. GigaScience, 6(4), gix002, 2017.
21. Zhang, Z., Spectral and time-frequency analysis. In EEG Signal Processing and feature extraction (pp. 89-116). Springer, Singapore, 2019.
22. Li, M. W., Geng, J., Hong, W. C., Zhang, L. D., Periodogram estimation based on LSSVR-CCPSO compensation for forecasting ship motion. Nonlinear Dynamics, 97(4), 2579-2594, 2019.
23. Francis, M.N., Keran, M.P., Chetan, R., Krupa, B.N., EEG-controlled robot navigation using Hjorth parameters and Welch-psd. International Journal of Intelligent Engineering and Systems, 14(4), 231-240, 2021.
24. Jin, X., Wang, Y., Hong, W., Power spectrum estimation method based on Matlab. In Proceedings of the 3rd International Conference on Vision, Image and Signal Processing (pp. 1-5), 2019.
25. Welch, P., The use of fast Fourier transform for the estimation of power spectra: a method based on time averaging over short, modified periodograms. IEEE Transactions on Audio and Electroacoustics, 15(2), 70-73, 1967.
26. Wieczorek, M. A., Simons, F. J., Minimum-variance multitaper spectral estimation on the sphere. Journal of Fourier Analysis and Applications, 13(6), 665-692, 2007.
27. Güneç, K., Kasım, Ö., Tosun, M., Büyükköroğlu, E., Estimation of pain threshold from EEG signals of subjects in physical therapy using long-short-term memory deep learning model. Uludağ University Journal of The Faculty of Engineering, 26(2), 447-460, 2021.
28. Lashgari, E., Liang, D., Maoz, U., Data augmentation for deep-learning-based electroencephalography. Journal of Neuroscience Methods, 108885, 2020.
29. Sakai, A., Minoda, Y., Morikawa, K., Data augmentation methods for machine-learning-based classification of bio-signals. In 2017 10th Biomedical Engineering International Conference (BMEiCON) (pp. 1-4). IEEE, 2017.
30. Kourou, K., Exarchos, T. P., Exarchos, K. P., Karamouzis, M. V., Fotiadis, D. I., Machine learning applications in cancer prognosis and prediction. Computational and Structural Biotechnology Journal, 13, 8-17, 2015.
31. Kong, W., Dong, Z.Y., Jia, Y., Hill, D.J., Xu, Y., Zhang, Y., Short-term residential load forecasting based on LSTM recurrent neural network. IEEE Transactions on Smart Grid, 10(1), 841-851, 2017.
32. Alhagry, S., Fahmy, A.A., El-Khoribi, R.A., Emotion recognition based on EEG using LSTM recurrent neural network. International Journal of Advanced Computer Science and Applications, 8(10), 355-358, 2017.
33. Liu, G., Guo, J., Bidirectional LSTM with attention mechanism and convolutional layer for text classification. Neurocomputing, 337, 325-338, 2019.
34. Durmuş, G., Soğukpinar, İ., Makine öğrenmesi teknikleri ile ikili yürütülebilir dosyalarda arabellek taşması zayıflığı analizi için yeni bir yaklaşım. Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi, 34(4), 1695-1704, 2019.
35. Göker, H., Bülbül, H. I., Improving an early warning system to prediction of student examination achievement. In 2014 13th international conference on machine learning and applications (pp. 568-573). IEEE, 2014.
36. Zwicker, J. G., Missiuna, C., Harris, S. R., Boyd, L. A., Brain activation of children with developmental coordination disorder is different than peers. Pediatrics, 126(3), e678-e686, 2010.
37. Tosun, M., Çetin, O. A new phase-based feature extraction method for four-class motor imagery classification. Signal, Image and Video Processing, 16(1), 283-290, 2022)
38. Saeedi, A., Saeedi, M., Maghsoudi, A., Shalbaf, A., Major depressive disorder diagnosis based on effective connectivity in EEG signals: A convolutional neural network and long short-term memory approach. Cognitive Neurodynamics, 15(2), 239-252, 2021.
39. Singh, K., Malhotra, J., Two-layer LSTM network-based prediction of epileptic seizures using EEG spectral features. Complex & Intelligent Systems, 1-14, 2022.
40. Nikhil Chandran, A., Sreekumar, K., Subha, D. P., EEG-based automated detection of schizophrenia using long short-term memory (LSTM) network. In Advances in Machine Learning and Computational Intelligence (pp. 229-236). Springer, Singapore, 2021.