{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,5]],"date-time":"2026-05-05T21:50:57Z","timestamp":1778017857948,"version":"3.51.4"},"reference-count":34,"publisher":"MIT Press","issue":"4","license":[{"start":{"date-parts":[[2025,3,3]],"date-time":"2025-03-03T00:00:00Z","timestamp":1740960000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2025,3,18]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Forecasting electroencephalography (EEG) signals, that is, estimating future values of the time series based on the past ones, is essential in many real-time EEG-based applications, such as brain\u2013computer interfaces and closed-loop brain stimulation. As these applications are becoming more and more common, the importance of a good prediction model has increased. Previously, the autoregressive model (AR) has been employed for this task; however, its prediction accuracy tends to fade quickly as multiple steps are predicted. We aim to improve on this by applying probabilistic deep learning to make robust longer-range forecasts.<\/jats:p>\n                  <jats:p>For this, we applied the probabilistic deep neural network model WaveNet to forecast resting-state EEG in theta- (4\u20137.5 Hz) and alpha-frequency (8\u201313 Hz) bands and compared it to the AR model.<\/jats:p>\n                  <jats:p>WaveNet reliably predicted EEG signals in both theta and alpha frequencies 150 ms ahead, with mean absolute errors of 1.0 \u00b1 1.1 \u00b5V (theta) and 0.9 \u00b1 1.1 \u00b5V (alpha), and outperformed the AR model in estimating the signal amplitude and phase. Furthermore, we found that the probabilistic approach offers a way of forecasting even more accurately while effectively discarding uncertain predictions.<\/jats:p>\n                  <jats:p>We demonstrate for the first time that probabilistic deep learning can be used to forecast resting-state EEG time series. In the future, the developed model can enhance the real-time estimation of brain states in brain\u2013computer interfaces and brain stimulation protocols. It may also be useful for answering neuroscientific questions and for diagnostic purposes.<\/jats:p>","DOI":"10.1162\/neco_a_01743","type":"journal-article","created":{"date-parts":[[2025,3,3]],"date-time":"2025-03-03T16:32:40Z","timestamp":1741019560000},"page":"793-814","update-policy":"https:\/\/doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":11,"title":["Enhanced EEG Forecasting: A Probabilistic Deep Learning Approach"],"prefix":"10.1162","volume":"37","author":[{"given":"Hanna","family":"Pankka","sequence":"first","affiliation":[{"name":"Department of Neuroscience and Biomedical Engineering, Aalto University School of Science, FI-02150 Espoo, Finland hanna.e.pankka@aalto.fi"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jaakko","family":"Lehtinen","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Aalto University School of Science, FI-02150 Espoo, Finland"},{"name":"NVIDIA, FI-00180 Helsinki, Finland jaakko.lehtinen@aalto.fi"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Risto J.","family":"Ilmoniemi","sequence":"additional","affiliation":[{"name":"Department of Neuroscience and Biomedical Engineering, Aalto University School of Science, FI-02150 Espoo, Finland risto.ilmoniemi@aalto.fi"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Timo","family":"Roine","sequence":"additional","affiliation":[{"name":"Department of Neuroscience and Biomedical Engineering, Aalto University School of Science, FI-02150 Espoo, Finland timo.roine@aalto.fi"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"281","published-online":{"date-parts":[[2025,3,18]]},"reference":[{"key":"2025060214005286500_bib1","author":"Abadi","year":"2016","journal-title":"TensorFlow: Large-scale machine learning on heterogeneous distributed systems"},{"key":"2025060214005286500_bib2","doi-asserted-by":"publisher","DOI":"10.1007\/s42979-020-00345-2","article-title":"Analyzing intracranial EEG in pharmacoresistant epilepsy patients using hidden Markov models and time series forecasting methods","volume":"1","author":"Bhowmick","year":"2020","journal-title":"SN Computer Science"},{"key":"2025060214005286500_bib3","doi-asserted-by":"publisher","first-page":"159","DOI":"10.1007\/BF00243291","article-title":"Non-linear and linear forecasting of the EEG time series","volume":"66","author":"Blinowska","year":"1991","journal-title":"Biological Cybernetics"},{"key":"2025060214005286500_bib4","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1162\/CPSY_a_00024","article-title":"Multiple dissociations between comorbid depression and anxiety on reward and punishment processing: Evidence from computationally informed EEG","volume":"3","author":"Cavanagh","year":"2019","journal-title":"Computational Psychiatry"},{"issue":"3","key":"2025060214005286500_bib5","doi-asserted-by":"publisher","first-page":"753","DOI":"10.1109\/TBME.2011.2109715","article-title":"Real-time brain oscillation detection and phase-locked stimulation using autoregressive spectral estimation and time-series forward prediction","volume":"60","author":"Chen","year":"2013","journal-title":"IEEE Transactions on Biomedical Engineering"},{"issue":"4","key":"2025060214005286500_bib6","doi-asserted-by":"publisher","first-page":"461","DOI":"10.1109\/TNSRE.2005.857690","article-title":"A time-series prediction approach for feature extraction in a brain-computer interface","volume":"13","author":"Coyle","year":"2005","journal-title":"IEEE Transactions on Neural Systems and Rehabilitation Engineering"},{"key":"2025060214005286500_bib7","doi-asserted-by":"crossref","first-page":"2749","DOI":"10.1109\/IJCNN.2011.6033579","article-title":"Classification of EEG during imagined mental tasks by forecasting with Elman recurrent neural networks","volume-title":"Proceedings of the 2011 International Joint Conference on Neural Networks","author":"Forney","year":"2011"},{"key":"2025060214005286500_bib8","doi-asserted-by":"crossref","DOI":"10.3389\/fnhum.2014.00122","article-title":"Coupling brain-machine interfaces with cortical stimulation for brain-state dependent stimulation: Enhancing motor cortex excitability for neurorehabilitation","volume":"8","author":"Gharabaghi","year":"2014","journal-title":"Frontiers in Human Neuroscience"},{"key":"2025060214005286500_bib9","first-page":"1321","article-title":"On calibration of modern neural networks","volume":"70","author":"Guo","year":"2017","journal-title":"Proceedings of the 34th International Conference on Machine Learning"},{"issue":"4","key":"2025060214005286500_bib10","doi-asserted-by":"publisher","first-page":"261","DOI":"10.1080\/21646821.2016.1245563","article-title":"American Clinical Neurophysiology Society guideline 4: Recording clinical EEG on digital media","volume":"56","author":"Halford","year":"2016","journal-title":"Neurodiagnostic Journal"},{"issue":"4","key":"2025060214005286500_bib11","doi-asserted-by":"publisher","DOI":"10.1093\/sleep\/zsaa227","article-title":"Phase-locked auditory stimulation of theta oscillations during rapid eye movement sleep","volume":"44","author":"Harrington","year":"2021","journal-title":"Sleep"},{"issue":"2","key":"2025060214005286500_bib12","doi-asserted-by":"publisher","first-page":"295","DOI":"10.1016\/j.jneumeth.2010.03.030","article-title":"EEG-based motor imagery classification using neuro-fuzzy prediction and wavelet fractal features","volume":"189","author":"Hsu","year":"2010","journal-title":"Journal of Neuroscience Methods"},{"issue":"12","key":"2025060214005286500_bib13","doi-asserted-by":"publisher","DOI":"10.1093\/sleep\/zsae193","article-title":"Closed-loop auditory stimulation targeting alpha and theta oscillations during REM sleep induces phase-dependent power and frequency changes","volume":"47","author":"Jaramillo","year":"2024","journal-title":"Sleep"},{"key":"2025060214005286500_bib14","author":"Kingma","year":"2014","journal-title":"Adam: A method for stochastic optimization"},{"issue":"2s","key":"2025060214005286500_bib15","doi-asserted-by":"crossref","first-page":"575","DOI":"10.52783\/jes.1478","article-title":"Forecasting the future: A comprehensive review of time series prediction techniques","volume":"20","author":"Kolambe","year":"2024","journal-title":"Journal of Electrical Systems"},{"key":"2025060214005286500_bib16","doi-asserted-by":"publisher","first-page":"2551","DOI":"10.1007\/s00521-017-3213-3","article-title":"EEG classification using recurrent adaptive neuro-fuzzy network based on time-series prediction","volume":"31","author":"Komijani","year":"2019","journal-title":"Neural Computing and Applications"},{"key":"2025060214005286500_bib17","article-title":"DiffWave: A versatile diffusion model for audio synthesis","volume-title":"Proceedings of the International Conference on Learning Representations","author":"Kong","year":"2021"},{"issue":"31","key":"2025060214005286500_bib18","doi-asserted-by":"publisher","first-page":"10122","DOI":"10.1523\/JNEUROSCI.23-31-10122.2003","article-title":"Interplay of electroencephalogram phase and auditory-evoked neural activity","volume":"23","author":"Kruglikov","year":"2003","journal-title":"Journal of Neuroscience"},{"key":"2025060214005286500_bib19","doi-asserted-by":"publisher","DOI":"10.3389\/fnins.2017.00401","article-title":"A fast EEG forecasting algorithm for phase-locked transcranial electrical stimulation of the human brain","volume":"11","author":"Mansouri","year":"2017","journal-title":"Frontiers in Neuroscience"},{"issue":"3","key":"2025060214005286500_bib20","doi-asserted-by":"publisher","DOI":"10.1088\/1741-2552\/ab8683","article-title":"Estimation of phase in EEG rhythms for real-time applications","volume":"17","author":"McIntosh","year":"2020","journal-title":"Journal of Neural Engineering"},{"key":"2025060214005286500_bib21","author":"Ping","year":"2018","journal-title":"ClariNet: Parallel wave generation in end-to-end text-to-speech"},{"key":"2025060214005286500_bib22","first-page":"3617","article-title":"WaveGlow: A flow-based generative network for speech synthesis","volume-title":"Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing","author":"Prenger","year":"2019"},{"key":"2025060214005286500_bib23","article-title":"Deep state space models for time series forecasting","volume-title":"Advances in neural information processing systems","author":"Rangapuram","year":"2018"},{"issue":"5","key":"2025060214005286500_bib24","doi-asserted-by":"publisher","DOI":"10.1088\/1741-2552\/ab260c","article-title":"Deep learning-based electroencephalography analysis: A systematic review","volume":"16","author":"Roy","year":"2019","journal-title":"Journal of Neural Engineering"},{"issue":"3","key":"2025060214005286500_bib25","doi-asserted-by":"publisher","first-page":"1181","DOI":"10.1016\/j.ijforecast.2019.07.001","article-title":"DeepAR: Probabilistic forecasting with autoregressive recurrent networks","volume":"36","author":"Salinas","year":"2020","journal-title":"International Journal of Forecasting"},{"issue":"10","key":"2025060214005286500_bib26","doi-asserted-by":"publisher","DOI":"10.3390\/app10103616","article-title":"Time-series prediction of the oscillatory phase of EEG signals using the least mean square algorithm-based AR model","volume":"10","author":"Shakeel","year":"2020","journal-title":"Applied Sciences"},{"key":"2025060214005286500_bib27","article-title":"Brain state-dependent brain stimulation with real-time electroencephalography: Triggered transcranial magnetic stimulation","volume":"150","author":"Stefanou","year":"2019","journal-title":"JoVE (Journal of Visualized Experiments)"},{"key":"2025060214005286500_bib28","author":"van den Oord","year":"2016","journal-title":"WaveNet: A generative model for raw audio"},{"key":"2025060214005286500_bib29","first-page":"3918","article-title":"Parallel WaveNet: Fast high-fidelity speech synthesis","volume-title":"Proceedings of the 35th International Conference on Machine Learning","author":"van den Oord","year":"2018"},{"issue":"11\u201312","key":"2025060214005286500_bib30","doi-asserted-by":"crossref","first-page":"3224","DOI":"10.1111\/ejn.14931","article-title":"Can the occipital alpha-phase speed up visual detection through a real-time EEG-based brain\u2013computer interface (BCI)?","volume":"55","author":"Vig\u00fce-Guix","year":"2022","journal-title":"European Journal of Neuroscience"},{"issue":"818","key":"2025060214005286500_bib31","first-page":"518","article-title":"On periodicity in series of related terms","volume":"131","author":"Walker","year":"1931","journal-title":"Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character"},{"key":"2025060214005286500_bib32","doi-asserted-by":"publisher","first-page":"267","DOI":"10.1098\/rsta.1927.0007","article-title":"On a method of investigating periodicities in disturbed series, with special reference to Wolfer\u2019s sunspot numbers","volume":"226","author":"Yule","year":"1927","journal-title":"Philosophical Transactions of the Royal Society of London. Series A"},{"issue":"2","key":"2025060214005286500_bib33","doi-asserted-by":"publisher","first-page":"374","DOI":"10.1016\/j.brs.2017.11.016","article-title":"Real-time EEG-defined excitability states determine efficacy of TMS-induced plasticity in human motor cortex","volume":"11","author":"Zrenner","year":"2018","journal-title":"Brain Stimulation"},{"key":"2025060214005286500_bib34","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2020.116761","article-title":"The shaky ground truth of real-time phase estimation","volume":"214","author":"Zrenner","year":"2020","journal-title":"NeuroImage"}],"container-title":["Neural Computation"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/direct.mit.edu\/neco\/article-pdf\/37\/4\/793\/2506381\/neco_a_01743.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/direct.mit.edu\/neco\/article-pdf\/37\/4\/793\/2506381\/neco_a_01743.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,2]],"date-time":"2025-06-02T14:01:39Z","timestamp":1748872899000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/neco\/article\/37\/4\/793\/128199\/Enhanced-EEG-Forecasting-A-Probabilistic-Deep"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,3,18]]},"references-count":34,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2025,3,18]]},"published-print":{"date-parts":[[2025,3,18]]}},"URL":"https:\/\/doi.org\/10.1162\/neco_a_01743","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2024.01.16.575836","asserted-by":"object"}]},"ISSN":["0899-7667","1530-888X"],"issn-type":[{"value":"0899-7667","type":"print"},{"value":"1530-888X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2025,4]]},"published":{"date-parts":[[2025,3,18]]}}}