Neural Noise

BCI closed-loop systems for neurorehabilitation: AI and machine learning innovations

2 min read

  • BCI closed-loop systems are a promising tool in healthcare and wellness monitoring, particularly in neurorehabilitation and cognitive assessment.1
  • There is a critical need for real-time, non-invasive monitoring technologies for neurological disorders including Alzheimer’s disease and related dementias.1
  • BCIs enable direct communication between the brain and external devices, leveraging artificial intelligence.1
  • A systematic review in JMIR Biomedical Engineering synthesizes AI and machine learning innovations in BCI closed-loop systems for neurorehabilitation and brain–device communication.1 1

Weekly enrichment (2026-07-20)

  • Full title and venue: “Advancing Brain–Computer Interface Closed-Loop Systems for Neurorehabilitation: Systematic Review of AI and Machine Learning Innovations in Biomedical Engineering,” JMIR Biomedical Engineering, 5 November 2025, 10:e72218 (doi:10.2196/72218; PMID 41191851; PMCID PMC12588595).23
  • Design: a PRISMA-guided systematic review of studies published 2019–2024, searching PubMed, IEEE, ACM, and Scopus with predefined keywords spanning BCIs, AI/ML, and Alzheimer’s disease and related dementias (AD/ADRD).34
  • Screening funnel: 220 records were identified (43 PubMed, 22 IEEE, 114 ACM, 41 Scopus); after removing 8 duplicates, 212 were screened, 33 full texts were assessed, and 18 studies met the final inclusion criteria.4
  • Exclusion detail: screening removed records that were out of context (n=84), not relevant to the research questions (n=94), or inaccessible (n=1); full-text exclusions were theses/books (n=9), report articles (n=4), and poor-quality studies (n=2).4
  • Key ML methods identified: transfer learning (TL), support vector machines (SVMs), and convolutional neural networks (CNNs), which improved signal classification, feature extraction, and real-time adaptability for monitoring cognitive states.23
  • Reported barriers: long calibration sessions, high computational cost, data-security risks, and variability in neural signals across sessions and users.3
  • Proposed solutions: improved sensor technology, more efficient calibration protocols, and advanced AI-driven decoding models.23
  • Clinical implication: AI-integrated closed-loop BCIs could power real-time alert systems that assist caregivers in managing AD/ADRD patients, supporting non-invasive neurorehabilitation and cognitive assessment.23

Footnotes

  1. https://biomedeng.jmir.org/2025/1/e72218 ↩ ↩2 ↩3 ↩4 ↩5

  2. https://doi.org/10.2196/72218 ↩ ↩2 ↩3 ↩4

  3. https://europepmc.org/article/MED/41191851 ↩ ↩2 ↩3 ↩4 ↩5 ↩6

  4. https://biomedeng.jmir.org/2025/1/e72218/PDF ↩ ↩2 ↩3

Graph View