Authors: Aragón Daud, Agustina; Boulakis, Paradeisios Alexandros; Simos, Nikolaos-Ioannis; Balla, Marion; Raimondo, Federico; Demertzi, Athena
Contact: a.aragondaud@uliege.be
The MBDecoder project aim to develop a machine learning (ML) model capable of identifying a mental state known as "Mind Blanking" (MB), in which people feel their mind is empty or have nothing to report about their internal experience [1,2], based on brain activity, particularly on whole-brain dynamic functional connectivity (dFC).
Target Sample: N=50 healthy adults
🚧 Full dataset acquisition ongoing
Inclusion Criteria:
- Right-handed (Edinburgh Handedness Inventory)
- Age > 18 years
Exclusion Criteria:
- Standard MRI contraindications
- Signs of dementia (AD8 Screening)
- History of neurological/psychiatric disorders
Acquisition Protocol: MRI (3T) with simulatenous EEG
- Resting-state: 10-minute fMRI-EEG
- Experience Sampling Task: 40-minute fMRI-EEG with intermittent self-report of mental state (Thought, Mind Blanking, Asleep, Sensations) + arousal rating
Questionnaires:
- Mind Blanking Questionnaire (MBQ)
- Metacognitions Questionnaire-30 (MCQ-30)
- Attentional Control Scale (ACS)
- Epworth Sleepiness Scale (ESS)
- Mind Blaking Self-report Questionnaire (OSF link)
- Amsterdam Resting-state Questionnaire (ARQ)
- Ad-hoc questions
- Resting state acquisition:
experiment/run_rest.py - Experince Sampling Task:
experiment/run_task.pyConda Environment for PsychoPy experiment:experiment/psychopy_env.yml
- fMRIprep for fMRI preprocessing
- EEGLAB-Matlab for initial EEG fMRI-related artifacts cleaning
-
- MNE-Python for EEG processing
- Python 3.8+ with scientific computing stack (NumPy, SciPy, scikit-learn) 🚧 Conda Environment to be uploaded
- Convert raw data to BIDS format:
00_bids/ - Run fMRI and EEG preprocessing:
01_preprocessing🚧 Further analysis to be uploaded
1) Convert RAW to BIDS: bash src_to_bids.sh XXX YYYY (where XXX is subject ID in BIDS format and YYYY is MRI scanner id (e.g, 4540)
2) Preprocess fMRI: could be run as bash src_to_bids.sh XXX (where XXX is subject ID in BIDS format, line 49 uncommented) or bash src_to_bids.sh (if line 48 is uncommented and 49 is commented)
**3) Preprocess EEG:
(3.1) Clean MRI-related artifacts:bash run_eeg.sh --subject XXX XXX --task ES rest (it can run as many subjects as specificied)
(3.2) Convert to fif: python 1_convert_to_fif.py
(3.3) Pick bad channels (manual selection): bash 2_pick_bad_chs.sh /data/project/mb_decoder/data/bids/mb_decoder (manually select bad channels in pop-up window)
(3.4) Cut in epochs: python 3_cut_epochs.py
(3.5) Pick bad epochs (manual selection): bash 4_pick_bad_epochs.sh /data/project/mb_decoder/data/bids/mb_decoder (manually selected bad epochs in pop-up window)
(3.6) Run ICA: python 5_run_ica.py
(3.7) Create ICA report (manual selection): bash 6_create_ica_report.sh (open report and choose components to remove, then save the information in the .json)
(3.8) Apply ICA: python 8_apply_ica.py
-
Open Science Framework (OSF): https://osf.io/kxj6z/overview 🚧 To be updated
- Contains: Pilot data, study protocol, and preliminary materials
- Status: 🚧 Regularly updated as the project progresses
-
Data Repository: https://gitlab.uliege.be/poc/datasets/mb_decoder# 🚧 To be updated
- Contains: Full dataset in BIDS format
- Status: 🚧 Under active development
Data Access Policy:
- Pilot data is currently available on OSF
- Full dataset will be uploaded to the ULiège GitLab repository
- All data will be made publicly available upon publication of associated papers
This project is actively under development. Please contact the authors for the most current status.
[1] Andrillon, T., Lutz, A., Windt, J., & Demertzi, A. (2025). Where is my mind? A neurocognitive investigation of mind blanking. Trends in cognitive sciences, 29(7), 600-613. https://doi.org/10.1016/j.tics.2025.02.002
[2] Boulakis, P. A., & Demertzi, A. (2025). Relating mind-blanking to the content and dynamics of spontaneous thinking. Current Opinion in Behavioral Sciences, 61, 101481. https://doi.org/10.1016/j.cobeha.2024.101481
