pith. sign in

arxiv: 2509.10082 · v1 · submitted 2025-09-12 · 📡 eess.SP · cs.LG

FetalSleepNet: A Transfer Learning Framework with Spectral Equalisation Domain Adaptation for Fetal Sleep Stage Classification

Weitao Tang , Johann Vargas-Calixto , Nasim Katebi , Nhi Tran , Sharmony B. Kelly , Gari D. Clifford , Robert Galinsky , Faezeh Marzbanrad This is my paper

Pith reviewed 2026-05-18 18:02 UTC · model grok-4.3

classification 📡 eess.SP cs.LG
keywords fetal EEGsleep stage classificationtransfer learningdomain adaptationspectral equalisationovine modeldeep neural networkEEG signal processing
0
0 comments X

The pith

Transfer learning from adult EEG with spectral equalisation classifies fetal sheep sleep stages at 86.6 percent accuracy.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper introduces FetalSleepNet as the first deep learning approach to automated sleep stage classification from fetal EEG recorded in sheep. It starts from a lightweight network pretrained on adult EEG and applies full fine-tuning together with spectral equalisation to reduce differences between the adult and fetal signal domains. The combined method reaches 86.6 percent accuracy and a macro F1-score of 62.5, exceeding baseline models. Accurate fetal sleep staging could support earlier detection of disrupted brain maturation tied to pregnancy complications such as hypoxia. The compact design targets eventual use in low-power, real-time fetal monitoring devices.

Core claim

A lightweight deep neural network originally trained for adult EEG sleep staging can be transferred to ovine fetal EEG through full fine-tuning and spectral equalisation domain adaptation. This yields 86.6 percent accuracy and 62.5 macro F1-score on fetal sleep stage classification, outperforming baseline models and establishing the first published deep learning framework for the task.

What carries the argument

Spectral equalisation-based domain adaptation applied during full fine-tuning of an adult EEG sleep staging network to align with fetal ovine EEG signals.

If this is right

  • The model can serve as a label engine to generate large-scale weak or semi-supervised labels for training on less invasive signals such as Doppler ultrasound or electrocardiogram.
  • Its lightweight architecture supports deployment in low-power, real-time, and wearable fetal monitoring systems.
  • Automated classification may enable earlier identification of abnormal fetal brain maturation associated with complications like hypoxia or intrauterine growth restriction.
  • The approach demonstrates a practical route for adapting adult-derived EEG models to fetal data where manual interpretation is laborious.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same spectral equalisation step could be tested on human fetal EEG recordings to check whether the domain gap behaves similarly across species.
  • Performance gains point to potential use in multi-modal systems that combine EEG labels with Doppler or ECG data for comprehensive fetal monitoring.
  • The method may extend to other cross-age or cross-condition EEG transfer tasks where spectral mismatch is the main barrier.

Load-bearing premise

Spectral equalisation sufficiently reduces the mismatch between adult and fetal EEG frequency content to allow effective transfer of sleep staging knowledge via fine-tuning.

What would settle it

A controlled test showing that full fine-tuning without spectral equalisation produces no accuracy gain over direct transfer or baselines on the same fetal EEG data.

Figures

Figures reproduced from arXiv: 2509.10082 by Faezeh Marzbanrad, Gari D. Clifford, Johann Vargas-Calixto, Nasim Katebi, Nhi Tran, Robert Galinsky, Sharmony B. Kelly, Weitao Tang.

Figure 1
Figure 1. Figure 1: Representative physiological signals illustrating sleep states in fetal [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: [Fs/2] conv, 128, /[Fs/4] 8 max-pool, /8 0.5 dropout 8 conv, 128, /1 8 conv, 128, /1 8 conv, 128, /1 4 max-pool, /4 0.5 dropout 256 LSTM 256 LSTM hi−1, ci−1 hi , ci 0.5 dropout 3 softmax ybi xi (2-channels) ai hi Representation Learning Sequence Learning conv layer: [filter size conv, n filters, /stride] max-pool layer: [pool size max-pool, /stride] Trainable Layer (pretrained init.) Non-trainable Layer No… view at source ↗
Figure 3
Figure 3. Figure 3: Comparison of PSD profiles for fetal and adult EEG. (a) fEEG PSD [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Comparison of validation and test F1 scores over training epochs [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Per-epoch performance of the Intermediate sleep state across 120 [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
read the original abstract

Introduction: This study presents FetalSleepNet, the first published deep learning approach to classifying sleep states from the ovine electroencephalogram (EEG). Fetal EEG is complex to acquire and difficult and laborious to interpret consistently. However, accurate sleep stage classification may aid in the early detection of abnormal brain maturation associated with pregnancy complications (e.g. hypoxia or intrauterine growth restriction). Methods: EEG electrodes were secured onto the ovine dura over the parietal cortices of 24 late gestation fetal sheep. A lightweight deep neural network originally developed for adult EEG sleep staging was trained on the ovine EEG using transfer learning from adult EEG. A spectral equalisation-based domain adaptation strategy was used to reduce cross-domain mismatch. Results: We demonstrated that while direct transfer performed poorly, full fine tuning combined with spectral equalisation achieved the best overall performance (accuracy: 86.6 percent, macro F1-score: 62.5), outperforming baseline models. Conclusions: To the best of our knowledge, FetalSleepNet is the first deep learning framework specifically developed for automated sleep staging from the fetal EEG. Beyond the laboratory, the EEG-based sleep stage classifier functions as a label engine, enabling large scale weak/semi supervised labeling and distillation to facilitate training on less invasive signals that can be acquired in the clinic, such as Doppler Ultrasound or electrocardiogram data. FetalSleepNet's lightweight design makes it well suited for deployment in low power, real time, and wearable fetal monitoring systems.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. This manuscript presents FetalSleepNet, the first deep learning framework for sleep stage classification from ovine fetal EEG. It uses transfer learning from an adult EEG model with a spectral equalisation domain adaptation technique to mitigate domain mismatch. The authors show that full fine-tuning with spectral equalisation yields the highest performance (accuracy 86.6%, macro F1-score 62.5), surpassing direct transfer and baselines. Potential applications to clinical monitoring via less invasive signals are discussed.

Significance. If the findings are robustly validated, the work could advance automated fetal brain activity analysis, facilitating early detection of maturation abnormalities. The transfer learning and lightweight architecture offer practical advantages for deployment in wearable systems. The novelty as the first such DL approach is a strength, though the lack of detailed ablations and experimental protocols reduces the immediate significance.

major comments (2)
  1. Results: The assertion that full fine tuning combined with spectral equalisation achieved the best performance (86.6% accuracy, 62.5 macro F1) relies on the assumption that spectral equalisation meaningfully reduces the adult-to-fetal domain shift. However, no quantitative domain-shift metrics (such as MMD or CORAL) before and after equalisation, nor an ablation study isolating the equalisation from plain fine-tuning on the fetal data, are provided. This is load-bearing for the central claim regarding the domain adaptation strategy.
  2. Methods/Experimental setup: Critical details are missing, including the number of sleep stages classified, how the 24 fetal subjects were split for training/testing, the cross-validation strategy, and statistical significance testing of the reported metrics. These omissions make it difficult to verify the robustness of the 86.6% accuracy and 62.5 macro F1 results.
minor comments (1)
  1. Abstract: The abstract refers to 'baseline models' without specifying them; adding this detail would improve clarity and allow better assessment of the performance gains.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed feedback on our manuscript. We have addressed each major comment below and will incorporate revisions to strengthen the presentation of our methods and results.

read point-by-point responses

  1. Referee: Results: The assertion that full fine tuning combined with spectral equalisation achieved the best performance (86.6% accuracy, 62.5 macro F1) relies on the assumption that spectral equalisation meaningfully reduces the adult-to-fetal domain shift. However, no quantitative domain-shift metrics (such as MMD or CORAL) before and after equalisation, nor an ablation study isolating the equalisation from plain fine-tuning on the fetal data, are provided. This is load-bearing for the central claim regarding the domain adaptation strategy.

    Authors: We agree that additional quantitative evidence would more directly support the contribution of spectral equalisation. While our results demonstrate that full fine-tuning with spectral equalisation outperforms direct transfer and other baselines, we did not report explicit domain discrepancy metrics or a dedicated ablation isolating the equalisation component. In the revised manuscript we will add these elements, including computation of metrics such as MMD on the learned features before and after equalisation together with an ablation table comparing fine-tuning alone versus fine-tuning plus spectral equalisation. revision: yes

  2. Referee: Methods/Experimental setup: Critical details are missing, including the number of sleep stages classified, how the 24 fetal subjects were split for training/testing, the cross-validation strategy, and statistical significance testing of the reported metrics. These omissions make it difficult to verify the robustness of the 86.6% accuracy and 62.5 macro F1 results.

    Authors: We acknowledge that these experimental details were not described with sufficient clarity. The revised manuscript will explicitly state the number of sleep stages, the precise subject-wise training/testing split of the 24 fetal recordings, the cross-validation procedure employed, and the statistical tests (with p-values) used to compare performance metrics against baselines. These additions will be placed in the Methods and Results sections to ensure full reproducibility and allow independent verification of the reported figures. revision: yes

Circularity Check

0 steps flagged

No circularity detected in empirical transfer learning and domain adaptation pipeline

full rationale

The paper reports an empirical machine-learning study: a lightweight adult-EEG sleep-staging network is transferred to a 24-subject ovine fetal EEG dataset, with a spectral-equalisation preprocessing step applied to reduce domain mismatch, followed by full fine-tuning. The headline performance numbers (86.6 % accuracy, 62.5 macro-F1) are obtained by direct evaluation on held-out fetal recordings. No equations, fitted parameters, or self-citations are presented that would reduce these measured outcomes to the inputs by construction, nor is any uniqueness theorem or ansatz smuggled in via prior work by the same authors. The derivation chain is therefore self-contained experimental reporting rather than a closed logical loop.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The framework depends on standard assumptions in transfer learning and domain adaptation for EEG signals, with the effectiveness of spectral equalisation being key but empirically tested only within this study.

free parameters (1)
  • Domain adaptation parameters
    Parameters involved in spectral equalisation to match adult and fetal EEG spectra, chosen or optimized during the process.
axioms (1)
  • domain assumption Fetal ovine EEG shares transferable features with adult EEG that can be aligned via spectral equalisation.
    This underpins the transfer learning and adaptation strategy described in the methods.

pith-pipeline@v0.9.0 · 5839 in / 1300 out tokens · 75445 ms · 2026-05-18T18:02:08.590361+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

48 extracted references · 48 canonical work pages

  1. [1]

    Human fetal heart rate variability- characteristics of autonomic regulation in the third trimester of gesta- tion,

    U. Schneider, B. Frank, A. Fiedler, C. Kaehler, D. Hoyer, M. Liehr, J. Haueisen, and E. Schleussner, “Human fetal heart rate variability- characteristics of autonomic regulation in the third trimester of gesta- tion,”Journal of Perinatal Medicine, vol. 36, no. 5, pp. 433–441, 2008

    work page 2008

  2. [2]

    Altered autonomic control of heart rate variability in the chronically hypoxic fetus,

    C. Shaw, B. Allison, N. Itani, K. Botting, Y . Niu, C. Lees, and D. Giussani, “Altered autonomic control of heart rate variability in the chronically hypoxic fetus,”The Journal of Physiology, vol. 596, no. 23, pp. 6105–6119, 2018

    work page 2018

  3. [3]

    Behavioural state transitions in healthy and growth retarded fetuses,

    D. Arduini, G. Rizzo, L. Caforio, M. R. Boccolini, C. Romanini, and S. Mancuso, “Behavioural state transitions in healthy and growth retarded fetuses,”Early Human Development, vol. 19, no. 3, pp. 155– 165, 1989

    work page 1989

  4. [4]

    Behavioural states in growth-retarded human fetuses,

    M. Van Vliet, C. Martin Jr, J. Nijhuis, and H. Prechtl, “Behavioural states in growth-retarded human fetuses,”Early Human Development, vol. 12, no. 2, pp. 183–197, 1985

    work page 1985

  5. [5]

    Investigation of nonlinear ECoG changes during spontaneous sleep state changes and cortical arousal in fetal sheep,

    M. Schwab, K. Schmidt, H. Witte, and R. M. Abrams, “Investigation of nonlinear ECoG changes during spontaneous sleep state changes and cortical arousal in fetal sheep,”Cerebral Cortex, vol. 10, no. 2, pp. 142–148, 2000

    work page 2000

  6. [6]

    Prostaglandin d synthase in the prenatal ovine brain and effects of its inhibition with selenium chloride on fetal sleep/wake activity in utero,

    B. Lee, J. J. Hirst, and D. W. Walker, “Prostaglandin d synthase in the prenatal ovine brain and effects of its inhibition with selenium chloride on fetal sleep/wake activity in utero,”Journal of Neuroscience, vol. 22, no. 13, pp. 5679–5686, 2002

    work page 2002

  7. [7]

    The instrumented fetal sheep as a model of cerebral white matter injury in the premature infant,

    S. A. Back, A. Riddle, J. Dean, and A. R. Hohimer, “The instrumented fetal sheep as a model of cerebral white matter injury in the premature infant,”Neurotherapeutics, vol. 9, no. 2, pp. 359–370, 2012

    work page 2012

  8. [8]

    Advancing fetal surveillance with physiological sensing: Detecting hypoxia in fetal sheep,

    W. Tang, N. Tran, N. Katebi, R. Sameni, G. D. Clifford, D. Walker, V . Horlali, C. Taylor, R. Galinsky, and F. Marzbanrad, “Advancing fetal surveillance with physiological sensing: Detecting hypoxia in fetal sheep,” in2024 IEEE SENSORS, 2024, pp. 1–4

    work page 2024

  9. [9]

    Are there behavioural states in the human fetus?

    J. Nijhuis, H. F. Prechtl, C. B. Martin Jr, and R. Bots, “Are there behavioural states in the human fetus?”Early Human Development, vol. 6, no. 2, pp. 177–195, 1982

    work page 1982

  10. [10]

    Prenatal development of sleep-wake patterns in sheep,

    H. H. Szeto and D. J. Hinman, “Prenatal development of sleep-wake patterns in sheep,”Sleep, vol. 8, no. 4, pp. 347–355, 1985

    work page 1985

  11. [11]

    Fetal sleep: A cross-species review of physiology, measurement, and classification,

    W. Tang, J. Vargas-Calixto, N. Katebi, R. Galinsky, G. D. Clifford, and F. Marzbanrad, “Fetal sleep: A cross-species review of physiology, measurement, and classification,” 2025, arXiv preprint arXiv:2506.21828. [Online]. Available: https://arxiv.org/abs/2506.21828

    work page arXiv 2025

  12. [12]

    Systematic trends across the night in human sleep cycles,

    I. Feinberg and T. Floyd, “Systematic trends across the night in human sleep cycles,”Psychophysiology, vol. 16, no. 3, pp. 283–291, 1979

    work page 1979

  13. [13]

    Quantifying the interactions between maternal and fetal heart rates by transfer entropy,

    F. Marzbanrad, Y . Kimura, M. Palaniswami, and A. H. Khandoker, “Quantifying the interactions between maternal and fetal heart rates by transfer entropy,”PloS One, vol. 10, no. 12, p. e0145672, 2015

    work page 2015

  14. [14]

    A computer-aided approach to detect the fetal behavioral states using multi-sensor mag- netocardiographic recordings,

    S. Vairavan, U. D. Ulusar, H. Eswaran, H. Preissl, J. D. Wilson, S. Mckelvey, C. L. Lowery, and R. B. Govindan, “A computer-aided approach to detect the fetal behavioral states using multi-sensor mag- netocardiographic recordings,”Computers in Biology and Medicine, vol. 69, pp. 44–51, 2016

    work page 2016

  15. [15]

    Evaluation of parameters for fetal behavioural state classification,

    L. Semeia, K. Sippel, J. Moser, and H. Preissl, “Evaluation of parameters for fetal behavioural state classification,”Scientific Reports, vol. 12, no. 1, p. 3410, 2022

    work page 2022

  16. [16]

    A quantitative method for classification of EEG in the fetal baboon,

    M. M. Myers, R. I. Stark, W. P. Fifer, P. G. Grieve, J. Haiken, K. Leung, and K. F. Schulze, “A quantitative method for classification of EEG in the fetal baboon,”American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, vol. 265, no. 3, pp. R706– R714, 1993

    work page 1993

  17. [17]

    Behavioral states in the fetal baboon,

    P. G. Grieve, M. M. Myers, and R. I. Stark, “Behavioral states in the fetal baboon,”Early Human Development, vol. 39, no. 3, pp. 159–175, 1994

    work page 1994

  18. [18]

    Classification of fetal behavioral states by using 1d-cnn based on fetal electrocardiography,

    A. Samjeed, M. Wahbah, L. Hadjileontiadis, and A. H. Khandoker, “Classification of fetal behavioral states by using 1d-cnn based on fetal electrocardiography,” in2022 Computing in Cardiology (CinC), vol. 498. IEEE, 2022, pp. 1–4

    work page 2022

  19. [19]

    Hidden markov models and deep neural networks: Seg- mentation of the fetal heart rate signal into behavioral patterns,

    L. Subitoni, “Hidden markov models and deep neural networks: Seg- mentation of the fetal heart rate signal into behavioral patterns,” Master’s thesis, Politecnico di Milano, Milan, Italy, 2023

    work page 2023

  20. [20]

    FetalSleepNet: A Transfer Learning Framework with Spectral Equalisation Domain Adaptation for Fetal Sleep Stage Classification,

    W. Tang, J. Vargas-Calixto, N. Katebi, S. B. Kelly, R. Galinsky, G. D. Clifford, and F. Marzbanrad, “FetalSleepNet: A Transfer Learning Framework with Spectral Equalisation Domain Adaptation for Fetal Sleep Stage Classification,” Aug. 2025. [Online]. Available: https://github.com/faemrad/FetalSleepNet

    work page 2025

  21. [21]

    The arrive guidelines 2.0: Updated guidelines for reporting animal research,

    N. Percie du Sert, V . Hurst, A. Ahluwalia, S. Alam, M. T. Avey, M. Baker, W. J. Browne, A. Clark, I. C. Cuthill, U. Dirnaglet al., “The arrive guidelines 2.0: Updated guidelines for reporting animal research,” Journal of Cerebral Blood Flow & Metabolism, vol. 40, no. 9, pp. 1769– 1777, 2020

    work page 2020

  22. [22]

    Prophylactic fetal creatine supplementation improves post-asphyxial eeg recovery and reduces seizures in fetal sheep: implications for hypoxic–ischemic encephalopathy,

    N. T. Tran, S. J. Ellery, S. B. Kelly, J. S ´evigny, M. Chatton, H. Lu, G. R. Polglase, R. J. Snow, D. W. Walker, and R. Galinsky, “Prophylactic fetal creatine supplementation improves post-asphyxial eeg recovery and reduces seizures in fetal sheep: implications for hypoxic–ischemic encephalopathy,”Annals of Neurology, vol. 97, no. 4, pp. 673–687, 2025

    work page 2025

  23. [23]

    Analysis of a sleep-dependent neuronal feedback loop: The slow-wave microcontinuity of the EEG,

    B. Kemp, A. H. Zwinderman, B. Tuk, H. A. Kamphuisen, and J. J. Oberye, “Analysis of a sleep-dependent neuronal feedback loop: The slow-wave microcontinuity of the EEG,”IEEE Transactions on Biomed- ical Engineering, vol. 47, no. 9, pp. 1185–1194, 2000

    work page 2000

  24. [24]

    Physiobank, physiotoolkit, and physionet: Components of a 13 new research resource for complex physiologic signals,

    A. L. Goldberger, L. A. Amaral, L. Glass, J. M. Hausdorff, P. C. Ivanov, R. G. Mark, J. E. Mietus, G. B. Moody, C.-K. Peng, and H. E. Stanley, “Physiobank, physiotoolkit, and physionet: Components of a 13 new research resource for complex physiologic signals,”Circulation, vol. 101, no. 23, pp. e215–e220, 2000

    work page 2000

  25. [25]

    Rechtschaffen and A

    A. Rechtschaffen and A. Kales, Eds.,A Manual of Standardized Ter- minology, Techniques and Scoring System for Sleep Stages of Human Subjects. Bethesda, Md: National Institute of Neurological Diseases and Blindness, Neurological Information Network, 1968, NIH Publication No. 204

    work page 1968

  26. [26]

    Be- havioural state linkage in the ovine fetus near term,

    N. Rao, A. Keen, M. Czikk, M. Frasch, and B. S. Richardson, “Be- havioural state linkage in the ovine fetus near term,”Brain Research, vol. 1250, pp. 149–156, 2009

    work page 2009

  27. [27]

    Real-time spectral analysis of the fetal EEG: A new approach to monitoring sleep states and fetal condition during labor,

    I. Thaler, R. Boldes, and I. Timor-Tritsch, “Real-time spectral analysis of the fetal EEG: A new approach to monitoring sleep states and fetal condition during labor,”Pediatric Research, vol. 48, no. 3, pp. 340–345, 2000

    work page 2000

  28. [28]

    Connexin hemichannel blockade improves survival of striatal gaba-ergic neurons after global cerebral ischaemia in term- equivalent fetal sheep,

    R. Galinsky, J. O. Davidson, C. A. Lear, L. Bennet, C. R. Green, and A. J. Gunn, “Connexin hemichannel blockade improves survival of striatal gaba-ergic neurons after global cerebral ischaemia in term- equivalent fetal sheep,”Scientific Reports, vol. 7, no. 1, p. 6304, 2017

    work page 2017

  29. [29]

    Magnesium sulphate reduces tertiary gliosis but does not improve eeg recovery or white or grey matter cell survival after asphyxia in preterm fetal sheep,

    R. Galinsky, S. K. Dhillon, S. B. Kelly, G. Wassink, J. O. Davidson, C. A. Lear, L. G. van den Heuij, L. Bennet, and A. J. Gunn, “Magnesium sulphate reduces tertiary gliosis but does not improve eeg recovery or white or grey matter cell survival after asphyxia in preterm fetal sheep,” The Journal of Physiology, vol. 601, no. 10, pp. 1999–2016, 2023

    work page 1999

  30. [30]

    EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis,

    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

    work page 2004

  31. [31]

    C. Iber, S. Ancoli-Israel, A. L. C. Jr., and S. F. Quan,The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. Westchester, IL, USA: American Academy of Sleep Medicine, 2007

    work page 2007

  32. [32]

    EEG time and frequency domain analyses of primary insomnia,

    S. T.-B. Hamida, T. Penzel, and B. Ahmed, “EEG time and frequency domain analyses of primary insomnia,” in2015 37th Annual Interna- tional Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE, 2015, pp. 6206–6209

    work page 2015

  33. [33]

    EEG analysis based on time domain properties,

    B. Hjorth, “EEG analysis based on time domain properties,”Electroen- cephalography and Clinical Neurophysiology, vol. 29, no. 3, pp. 306– 310, 1970

    work page 1970

  34. [34]

    Progressive inflammation reduces high-frequency EEG activity and cortical dendritic arborisation in late gestation fetal sheep,

    S. B. Kelly, J. M. Dean, V . A. Zahra, I. Dudink, A. Thiel, G. R. Polglase, S. L. Miller, S. B. Hooper, L. Bennet, A. J. Gunnet al., “Progressive inflammation reduces high-frequency EEG activity and cortical dendritic arborisation in late gestation fetal sheep,”Journal of Neuroinflammation, vol. 20, no. 1, p. 124, 2023

    work page 2023

  35. [35]

    The use of fast fourier transform for the estimation of power spectra: A method based on time averaging over short, modified peri- odograms,

    P. Welch, “The use of fast fourier transform for the estimation of power spectra: A method based on time averaging over short, modified peri- odograms,”IEEE Transactions on Audio and Electroacoustics, vol. 15, no. 2, pp. 70–73, 2003

    work page 2003

  36. [36]

    Rhinal–hippocampal theta coherence during declarative memory for- mation: interaction with gamma synchronization?

    J. Fell, P. Klaver, H. Elfadil, C. Schaller, C. E. Elger, and G. Fern ´andez, “Rhinal–hippocampal theta coherence during declarative memory for- mation: interaction with gamma synchronization?”European Journal of Neuroscience, vol. 17, no. 5, pp. 1082–1088, 2003

    work page 2003

  37. [37]

    Tinysleepnet: An efficient deep learning model for sleep stage scoring based on raw single-channel EEG,

    A. Supratak and Y . Guo, “Tinysleepnet: An efficient deep learning model for sleep stage scoring based on raw single-channel EEG,” in2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC). IEEE, 2020, pp. 641–644

    work page 2020

  38. [38]

    Deepsleepnet: A model for automatic sleep stage scoring based on raw single-channel eeg,

    A. Supratak, H. Dong, C. Wu, and Y . Guo, “Deepsleepnet: A model for automatic sleep stage scoring based on raw single-channel eeg,”IEEE transactions on Neural Systems and Rehabilitation Engineering, vol. 25, no. 11, pp. 1998–2008, 2017

    work page 1998

  39. [39]

    Sleepeegnet: Automated sleep stage scoring with sequence to sequence deep learning approach,

    S. Mousavi, F. Afghah, and U. R. Acharya, “Sleepeegnet: Automated sleep stage scoring with sequence to sequence deep learning approach,” PloS One, vol. 14, no. 5, p. e0216456, 2019

    work page 2019

  40. [40]

    Xsleepnet: Multi-view sequential model for automatic sleep staging,

    H. Phan, O. Y . Ch ´en, M. C. Tran, P. Koch, A. Mertins, and M. De V os, “Xsleepnet: Multi-view sequential model for automatic sleep staging,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 9, pp. 5903–5915, 2021

    work page 2021

  41. [41]

    Sleeptransformer: Automatic sleep staging with interpretability and un- certainty quantification,

    H. Phan, K. Mikkelsen, O. Y . Ch´en, P. Koch, A. Mertins, and M. De V os, “Sleeptransformer: Automatic sleep staging with interpretability and un- certainty quantification,”IEEE Transactions on Biomedical Engineering, vol. 69, no. 8, pp. 2456–2467, 2022

    work page 2022

  42. [42]

    Spectral equalisation: A signal processing technique to increase pulse-echo signal bandwidth,

    L. van der Wal, A. Dijkhuizen, and G. Peels, “Spectral equalisation: A signal processing technique to increase pulse-echo signal bandwidth,” in Acoustical Imaging, M. Kavehet al., Eds. New York: Plenum Press, 1984, vol. 12, pp. 273–291

    work page 1984

  43. [43]

    Individual comparisons by ranking methods,

    F. Wilcoxon, “Individual comparisons by ranking methods,”Biometrics Bulletin, vol. 1, no. 6, pp. 80–83, 1945

    work page 1945

  44. [44]

    A simple sequentially rejective multiple test procedure,

    S. Holm, “A simple sequentially rejective multiple test procedure,” Scandinavian Journal of Statistics, pp. 65–70, 1979

    work page 1979

  45. [45]

    Sleep cycles and kinesis in the foetal lamb,

    Y . Ruckebusch, M. Gaujoux, and B. Eghbali, “Sleep cycles and kinesis in the foetal lamb,”Electroencephalography and Clinical Neurophysiology, vol. 42, no. 2, pp. 226–237, 1977

    work page 1977

  46. [46]

    Development of integrative autonomic nervous system function: An investigation based on time correlation in fetal heart rate patterns,

    U. Wallwitz, U. Schneider, S. Nowack, J. Feuker, S. Bauer, A. Rudolph, and D. Hoyer, “Development of integrative autonomic nervous system function: An investigation based on time correlation in fetal heart rate patterns,”Journal of Perinatal Medicine, vol. 40, no. 6, pp. 659–667, 2012

    work page 2012

  47. [47]

    Fetal autonomic brain age scores, segmented heart rate variability analysis, and traditional short term variability,

    D. Hoyer, E.-M. Kowalski, A. Schmidt, F. Tetschke, S. Nowack, A. Rudolph, U. Wallwitz, I. Kynass, F. Bode, J. Tegtmeyeret al., “Fetal autonomic brain age scores, segmented heart rate variability analysis, and traditional short term variability,”Frontiers in Human Neuroscience, vol. 8, p. 948, 2014

    work page 2014

  48. [48]

    Developmental milestones of the autonomic nervous system revealed via longitudinal monitoring of fetal heart rate variability,

    U. Schneider, F. Bode, A. Schmidt, S. Nowack, A. Rudolph, E.-M. Doelcker, P. Schlattmann, T. G ¨otz, and D. Hoyer, “Developmental milestones of the autonomic nervous system revealed via longitudinal monitoring of fetal heart rate variability,”PloS One, vol. 13, no. 7, p. e0200799, 2018

    work page 2018