A mobile EEG study of 32 healthy young adults walking overground and stepping over expected and unexpected obstacles projected onto the floor (Mustile et al., 2021) identified two oscillatory markers: an early frontal theta increase beginning as soon as an obstacle appears, reflecting proactive updating of the motor plan, and a post-crossing centro-parietal beta rebound reflecting reactive resetting of the motor system. The authors left open whether these markers describe ecologically valid and clinically usable phenomena: the paradigm used two-dimensional projected rather than real three-dimensional obstacles, cortical dynamics could not be linked to gait kinematics or to negotiation success, complex route geometry and inter-individual gait variability were untested, and clinical generalization was speculative. The literature since 2021 shows that generalization to Parkinson's disease has been demonstrated by the same group, and that EEG during overground walking with real foam obstacles exists in other paradigms, but no study has tested whether the proactive and reactive markers themselves — their engagement and timing — survive real three-dimensional obstacle negotiation in unconstrained walking, whether they correlate with gait parameters and success or failure, whether they generalize to older adults at fall risk and to dyspraxia, and whether marker-driven diagnostic monitoring or rehabilitation is effective. That surviving core is open.

Background

Electroencephalography (EEG) records the summed electrical activity of cortical neurons from electrodes on the scalp, and its oscillatory content is conventionally decomposed into frequency bands: theta (roughly 4–7 Hz), associated with cognitive control and action monitoring over frontal midline sites, and beta (roughly 13–30 Hz), associated with sensorimotor cortex status, where power decreases during movement and transiently increases above baseline after movement ends — the post-movement beta rebound, widely interpreted as a resetting or idling of the sensorimotor system. Cognitive control research distinguishes proactive control, in which the motor plan is adjusted in advance of an anticipated interference, from reactive control, which compensates after an unexpected event has occurred. Walking over obstacles is a natural test case for this distinction: an obstacle appearing in the walking path demands either anticipatory gait adjustment or online correction, followed by re-stabilization of the locomotor rhythm.

Traditional EEG requires participants to remain still, which is incompatible with walking. Mobile EEG — lightweight, wireless amplifier systems worn on the body or in a backpack — removes that constraint, at the cost of fewer electrodes (typically 32 rather than the 64–256 of laboratory systems), more movement and muscle artefact, and consequently no reliable source localization (identifying which brain regions generated the scalp signal). Overground studies before 2021 had already shown that gait-cycle-synchronized sensor-level EEG analysis outside the laboratory is feasible (Oliveira, Arguissain and Andersen, 2018, stepping-precision manipulation during overground walking) and that obstacle stepping can be studied at the source level with a dual-layer mobile EEG system validated on a head phantom, though only on a treadmill (Nordin, Hairston and Ferris, 2019, who found a supplementary-motor/premotor to posterior-parietal activation sequence after unexpected obstacle appearance).

Mustile and colleagues (2021) brought this logic into a real-world setting: 32 healthy young adults walked a long path on the floor of a large room while expected and unexpected obstacles were projected onto the floor in front of them, and 32-channel mobile EEG was time–frequency-analysed around obstacle events. Obstacle appearance, whether expected or not, elicited an early frontal theta increase beginning as soon as the obstacle appeared rather than when it was reached — arbitrating between an early-selection and a late-adaptation account of proactive control in favor of the former — and a clear centro-parietal beta rebound followed obstacle crossing regardless of whether the motor plan had required updating, interpreted as reactive resetting of the motor system.

In the discussion of that paper the authors laid out what remains unknown. The 32-channel montage precludes source localization and extracts fewer independent components (the artefact-rejection and source-separation workhorse of EEG analysis) than laboratory systems; obstacles were two-dimensional light projections rather than physical objects, which could interfere with natural leg mechanics; no gait measurement was synchronized with the EEG, so cortical dynamics could not be linked to heel-strike timing, gait kinematics, or the success or failure of negotiation; route geometry was simple, without repeated turns; and the markers had not been tested in the clinical populations — older adults, people with Parkinson's disease, people with dyspraxia (developmental coordination disorder) — for which they were proposed as monitoring and rehabilitation tools, via neurofeedback or non-invasive brain stimulation.

Since 2021 the field has moved along several of these axes: the same group repeated the projected-obstacle paradigm in Parkinson's disease (Mustile et al., 2023); a Brazilian group recorded 64-channel mobile EEG during overground walking over real foam obstacles in Parkinson's patients and older adults (Vitório et al., 2023) and combined fNIRS (functional near-infrared spectroscopy, a wearable optical measure of cortical haemodynamics) with sparse EEG during real obstacle avoidance with gait measurement (Orcioli-Silva et al., 2021); virtual and augmented-reality obstacle paradigms have been developed on treadmills and in overground settings (Lu et al., 2019; Bühler et al., 2026; Krugliak and Clarke, 2022; Krelove et al., 2025); and reviews have consolidated what mobile neuroimaging of walking has established and which methodological problems — notably the absence of standardized artefact-rejection pipelines — remain (Richer, Bradford and Ferris, 2024; Vinod et al., 2026). None of these, however, has tested the source study's specific proactive and reactive markers under ecologically valid obstacle negotiation.

Problem Statement

The source study established, in healthy young adults walking overground and stepping over expected and unexpected obstacles projected onto the floor, two oscillatory markers: a frontal theta increase beginning as soon as an upcoming obstacle appears (proactive control, supporting an early-selection account of motor-plan updating), and a post-crossing centro-parietal beta rebound (reactive resetting of the motor system). The open problem is the ecological and clinical validity of these markers — whether they describe obstacle negotiation as it actually occurs outside the laboratory and in the populations where it matters. Concretely, four coordinated questions jointly test this single objective: (1) Ecological validity: does negotiation of real three-dimensional obstacles during unconstrained overground walking, including route geometry with repeated turns, engage the same proactive frontal-theta and reactive beta-rebound processes, with the same timing, as the projected-obstacle paradigm — including the direct comparison of projected versus real obstacles that the source study called for? (2) Brain–behaviour links: with EEG synchronized to heel-strike timing and gait kinematics, do the timing and magnitude of the markers correlate with gait parameters and distinguish successful from failed obstacle negotiation, across inter-individual variability in walking speed and route execution? (3) Population validity: do the markers generalize to older adults at fall risk and to people with dyspraxia (developmental coordination disorder)? Generalization to Parkinson's disease is already demonstrated — with attenuated theta and beta modulation in patients — and is therefore not part of what remains to be shown, though finer clinical characterization is not excluded. (4) Translational validity: can the markers be used for diagnostic monitoring of cognitive-motor decline, or to drive rehabilitation interventions such as neurofeedback or non-invasive brain stimulation in patient groups? An answer showing that the markers do not survive real obstacles, gait-synchronized analysis, or clinical populations would be as substantive as one showing that they do.

References

  1. Magda Mustile, Dimitrios Kourtis, Simon Ladouce, Gemma Learmonth, Martin G. Edwards, David I. Donaldson, Magdalena Ietswaart (2021), Mobile EEG reveals functionally dissociable dynamic processes supporting real-world ambulatory obstacle avoidance: Evidence for early proactive control, European Journal of Neuroscience 54(12):8106–8119, DOI 10.1111/ejn.15120, https://doi.org/10.1111/ejn.15120
  2. Magda Mustile, Dimitrios Kourtis, Martin G. Edwards, Simon Ladouce, Daniele Volpe, Manuela Pilleri, Elisa Pelosin, Gemma Learmonth, David I. Donaldson, Magdalena Ietswaart (2023), Characterizing neurocognitive impairments in Parkinson's disease with mobile EEG when walking and stepping over obstacles, Brain Communications 5(6):fcad326, DOI 10.1093/braincomms/fcad326, https://doi.org/10.1093/braincomms/fcad326
  3. Andrew D. Nordin, W. David Hairston, Daniel P. Ferris (2019), Human electrocortical dynamics while stepping over obstacles, Scientific Reports 9:5311, DOI 10.1038/s41598-019-41131-2, https://doi.org/10.1038/s41598-019-41131-2
  4. Rodrigo Vitório, Ellen Lirani-Silva, Diego Orcioli-Silva, Victor Spiandor Beretta, Anderson Souza Oliveira, Lilian Teresa Bucken Gobbi (2023), Electrocortical Dynamics of Usual Walking and the Planning to Step over Obstacles in Parkinson's Disease, Sensors 23(10):4866, DOI 10.3390/s23104866, https://doi.org/10.3390/s23104866
  5. Diego Orcioli-Silva, Rodrigo Vitório, Priscila Nóbrega-Sousa, Victor Spiandor Beretta, Núbia Ribeiro da Conceição, Anderson Souza Oliveira, Marcelo Pinto Pereira, Lilian Teresa Bucken Gobbi (2021), Cortical Activity Underlying Gait Improvements Achieved With Dopaminergic Medication During Usual Walking and Obstacle Avoidance in Parkinson's Disease, Neurorehabilitation and Neural Repair 35(5):406–418, DOI 10.1177/15459683211000736, https://doi.org/10.1177/15459683211000736
  6. Chiahao Lu, Emily Twedell, Reem Elbasher, Michael McCabe, Colum D. MacKinnon, Scott E. Cooper (2019), Avoiding Virtual Obstacles During Treadmill Gait in Parkinson's Disease, Frontiers in Aging Neuroscience 11:76, DOI 10.3389/fnagi.2019.00076, https://doi.org/10.3389/fnagi.2019.00076
  7. Marco A. Bühler, Sylvain Baillet, Bradford J. McFadyen, Joyce Fung, Anouk Lamontagne (2026), The effects of time constraints on electrocortical dynamics underlying obstacle avoidance while walking, Cortex 201:81–93, DOI 10.1016/j.cortex.2026.04.017, https://doi.org/10.1016/j.cortex.2026.04.017
  8. Anderson Souza Oliveira, Federico Gabriel Arguissain, Ole Kæseler Andersen (2018), Cognitive Processing for Step Precision Increases Beta and Gamma Band Modulation During Overground Walking, Brain Topography 31(4):661–671, DOI 10.1007/s10548-018-0633-z, https://doi.org/10.1007/s10548-018-0633-z
  9. Alexandra Krugliak, Alex Clarke (2022), Towards real-world neuroscience using mobile EEG and augmented reality, Scientific Reports 12:6574, DOI 10.1038/s41598-022-06296-3, https://doi.org/10.1038/s41598-022-06296-3
  10. Natalie Richer, J. Cortney Bradford, Daniel P. Ferris (2024), Mobile neuroimaging: What we have learned about the neural control of human walking, with an emphasis on EEG-based research, Neuroscience & Biobehavioral Reviews 162:105718, DOI 10.1016/j.neubiorev.2024.105718, https://doi.org/10.1016/j.neubiorev.2024.105718
  11. Vaishali Vinod, Lara Johanna Papin, Robbin Romijnders, Walter Maetzler, Julius Welzel (2026), Preprocessing on the Go: Practices in Gait-Related Mobile EEG, Psychophysiology 63(6), DOI 10.1111/psyp.70352, https://doi.org/10.1111/psyp.70352
  12. Fares Al-Shargie, Michael Glassen, John DeLuca, Soha Saleh (2025), Brain Connectivity During Walking and Obstacle Avoidance in Persons With Multiple Sclerosis and Healthy Controls: A Pilot EEG Study, IEEE Transactions on Neural Systems and Rehabilitation Engineering 33:2945–2955, DOI 10.1109/TNSRE.2025.3592492, https://doi.org/10.1109/TNSRE.2025.3592492
  13. Samantha Marshall, Gianna Jeyarajan, Nicholas Hayhow, Raphael Gabiazon, Tia Seleem, Mathew R. Hammerstrom, Olav Krigolson, Lindsay S. Nagamatsu (2024), Cortical activation among young adults during mobility in an indoor real-world environment: a mobile EEG approach, Neuropsychologia 203:108971, DOI 10.1016/j.neuropsychologia.2024.108971, https://doi.org/10.1016/j.neuropsychologia.2024.108971
  14. L. Maureen Krelove, Anthony Machula, Lauren E. Sergio, George Mochizuki (2025), Prefrontal cortex activity during virtual obstacle avoidance and distracted walking: A methodological proof of concept using augmented reality and functional near-infrared spectroscopy, PLOS ONE 20(10):e0333622, DOI 10.1371/journal.pone.0333622, https://doi.org/10.1371/journal.pone.0333622