A Robust Multi-Branch CNN-LSTM Architecture for Cross-Subject Motor Imagery Classification.

Overview

Authors: Simone Zini¹, Federico Bidone¹, Paolo Napoletano¹

ORCID iDs: Simone Zini, Paolo Napoletano

Department of Informatics, Systems and Communication, University of Milano-Bicocca, Viale Sarca 336, 20126 Milano, Italy

Institutions: University of Milano-Bicocca (Italy)

Journal: Sensors (Basel, Switzerland), volume 26, issue 11, article 3310

Dates: received 31 March 2026; accepted 21 May 2026; published online 23 May 2026

Type: Research article · Language: English

License: CC BY

Identifiers: DOI 10.3390/s26113310 · PMID 42280831 · PMCID PMC13259246 · OpenAlex W7162264564

Open access: gold, a free copy (OpenAlex)

Status: data only

Categories: EEG (modality), human (organism)

Methods: Spectral & time-frequency, Preprocessing, Connectivity, Machine learning, Physiology & signal measures

Keywords: BCI, motor imagery, deep learning, LSTM, EEG

MeSH: Brain-Computer Interfaces*, Convolutional Neural Networks*, Imagination*, Long Short Term Memory*, Electroencephalography, Humans, Movement, Signal Processing, Computer-Assisted (* major topic)

Topic: EEG and Brain-Computer Interfaces (Cognitive Neuroscience, Neuroscience), according to OpenAlex

Funding: AdvaNced Technologies for Human-centrEd Medicine (ANTHEM) (PNC0000003)

Citations: not cited yet (Europe PMC); 24 references in the paper

Abstract

Brain–computer interfaces (BCIs) based on motor imagery (MI) aim to convert electroencephalographic (EEG) activity into reliable device commands across users and recording setups. However, low signal-to-noise ratio and strong inter-subject variability still limit true “plug-and-play” deployment without lengthy calibration. To address these challenges, we propose a multi-branch convolutional long short-term memory (CNN-LSTM) architecture that jointly performs multi-scale temporal feature extraction and within-trial sequence modeling. The model employs four parallel 1D convolutional branches with distinct kernel sizes, each followed by an LSTM module and late fusion, combined with group normalization and supervision over sequences of sub-windows within each trial. We evaluate the approach on the EEG Motor Movement/Imagery (EEGMMI) dataset from PhysioNet under strictly subject-independent conditions, and on the ISLab-MI Dataset, a 32-channel wearable-EEG collection designed to assess cross-setup robustness. On EEGMMI, the network achieves up to 82.63% accuracy for binary left/right MI and 74.10% for a four-class task using 4 s trials under 5-fold cross-validation, outperforming an EEGNet-style baseline by 1–10% depending on class count and window length. Under a leave-one-subject-out protocol, the model attains 74.9% mean accuracy for a three-class MI task. Zero-shot transfer to ISLab-MI yields 64.60% and 63.02% accuracy in three- and four-class settings, respectively, while brief subject-specific fine-tuning using only 20% of each session improves performance to 81.38% and 73.48%. These findings show that combining multi-scale convolutional feature extraction with explicit sequence modeling and robust normalization yields accurate, data-efficient, and portable MI decoders suitable for practical BCI applications.

Reproduced under the paper's license (CC BY), from the paper cited above.

Code

The paper links to its data, not to its authors' code: see the Data section.

The paper's code and data availability statement is in the Data section.

Tracing map

A tracing map links a paper to the code its authors published: this paper has none, so it has no map.

Data

Datasets cited

github.com/unimib-islab/eeg-bci-cross-subject-motor-imagery — at github.com; found in “Data Availability Statement”

Data Availability Statement

The public dataset supporting the results of this article is the EEG Motor Movement/Imagery (EEGMMI) dataset hosted on PhysioNet, and the authors confirm that the dataset is indicated in the reference list. The ISLab-MI dataset, collected by the authors, is available at https://github.com/unimib-islab/EEG-BCI-Cross-Subject-Motor-Imagery (accessed on 20 May 2026).

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 28 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 3 authors, 5 keywords, 8 MeSH terms, 1 funder, 22 references.

Cite

This paper

Zini, S., Bidone, F., & Napoletano, P. (2026). A Robust Multi-Branch CNN-LSTM Architecture for Cross-Subject Motor Imagery Classification. Sensors (Basel, Switzerland), 26(11), 3310. https://doi.org/10.3390/s26113310

BibTeX

@article{zini2026robust, author = {Zini, Simone and Bidone, Federico and Napoletano, Paolo}, title = {{A Robust Multi-Branch CNN-LSTM Architecture for Cross-Subject Motor Imagery Classification}}, journal = {Sensors (Basel, Switzerland)}, year = {2026}, month = may, volume = {26}, number = {11}, pages = {3310}, publisher = {Multidisciplinary Digital Publishing Institute (MDPI)}, issn = {1424-8220}, doi = {10.3390/s26113310}, url = {https://doi.org/10.3390/s26113310}, pmid = {42280831}, pmcid = {PMC13259246} }

RIS

TY - JOUR AU - Zini, Simone AU - Bidone, Federico AU - Napoletano, Paolo TI - A Robust Multi-Branch CNN-LSTM Architecture for Cross-Subject Motor Imagery Classification T2 - Sensors (Basel, Switzerland) J2 - Sensors (Basel) PY - 2026 DA - 2026/05/23 VL - 26 IS - 11 SP - 3310 SN - 1424-8220 PB - Multidisciplinary Digital Publishing Institute (MDPI) DO - 10.3390/s26113310 UR - https://doi.org/10.3390/s26113310 LA - en ER -

CSL-JSON

{ "id": "10.3390/s26113310", "type": "article-journal", "title": "A Robust Multi-Branch CNN-LSTM Architecture for Cross-Subject Motor Imagery Classification", "container-title": "Sensors (Basel, Switzerland)", "author": [ { "family": "Zini", "given": "Simone" }, { "family": "Bidone", "given": "Federico" }, { "family": "Napoletano", "given": "Paolo" } ], "container-title-short": "Sensors (Basel)", "volume": "26", "issue": "11", "page": "3310", "DOI": "10.3390/s26113310", "PMID": "42280831", "PMCID": "PMC13259246", "ISSN": "1424-8220", "publisher": "Multidisciplinary Digital Publishing Institute (MDPI)", "URL": "https://doi.org/10.3390/s26113310", "language": "en", "issued": { "date-parts": [ [ 2026, 5, 23 ] ] } }

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.