Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Real-Time Brain-Computer Interfaces

Tufvesson, Pex LU orcid (2025)
Abstract
A brain-computer interface (BCI) is a real-time system that converts a user’s brain activity into commands, enabling control over applications such as moving a cursor on the screen. This conversion is made possible by machine learning techniques and other algorithms.
Historically, BCI research has focused on the analysis of brain signals offline. However, increasing computational power and algorithmic sophistication now allow real-time interaction, enabling systems to interpret brain activity and translate it into commands instantaneously. This is critical for neuroprosthetic applications where even millisecond-level improvements can produce more natural movement. Even in more playful areas of use, such as brain-controlled computer... (More)
A brain-computer interface (BCI) is a real-time system that converts a user’s brain activity into commands, enabling control over applications such as moving a cursor on the screen. This conversion is made possible by machine learning techniques and other algorithms.
Historically, BCI research has focused on the analysis of brain signals offline. However, increasing computational power and algorithmic sophistication now allow real-time interaction, enabling systems to interpret brain activity and translate it into commands instantaneously. This is critical for neuroprosthetic applications where even millisecond-level improvements can produce more natural movement. Even in more playful areas of use, such as brain-controlled computer games, response time is important for creating an experience that can engage the user: "It’s no fun if it lags". A key driver of these advances is machine learning. Unlike earlier BCIs that relied on fixed models, machine learning allows systems to adapt and continuously learn from user signals and behavior, fostering a more intuitive human-machine interface through ongoing feedback and mutual learning.
Despite these promising developments, significant challenges remain: real-time BCI systems demand high-performance hardware and software, require efficient wireless communication, robust processing capabilities, and adaptive algorithms. Furthermore, integrating these systems into real-world applications raises questions about usability and safety.
This thesis addresses some of these challenges, particularly in the context of reducing cognitive load, addressing latency uncertainties and developing a modular BCI research framework. This work presents new insights and pathways toward functional BCI systems. Through experimentation and novel system architecture, this research lays a foundation for the future of BCI technology. (Less)
Please use this url to cite or link to this publication:
author
supervisor
opponent
  • Associate Professor Bougrain, Laurent, University of Lorraine, France
organization
publishing date
type
Thesis
publication status
published
subject
pages
225 pages
publisher
Department of Automatic Control, Lund University
defense location
Lecture Hall M:A, building M, Ole Römers väg 1F, Faculty of Engineering LTH, Lund University, Lund.
defense date
2025-11-21 09:15:00
ISBN
978-91-8104-695-3
978-91-8104-696-0
project
Real-Time Brain-Computer Interfaces
language
English
LU publication?
yes
id
d4f3e505-c282-4ab7-99ae-e14a5742380f
date added to LUP
2025-10-23 16:15:08
date last changed
2025-10-28 09:48:42
@phdthesis{d4f3e505-c282-4ab7-99ae-e14a5742380f,
  abstract     = {{A brain-computer interface (BCI) is a real-time system that converts a user’s brain activity into commands, enabling control over applications such as moving a cursor on the screen. This conversion is made possible by machine learning techniques and other algorithms.<br/> Historically, BCI research has focused on the analysis of brain signals offline. However, increasing computational power and algorithmic sophistication now allow real-time interaction, enabling systems to interpret brain activity and translate it into commands instantaneously. This is critical for neuroprosthetic applications where even millisecond-level improvements can produce more natural movement. Even in more playful areas of use, such as brain-controlled computer games, response time is important for creating an experience that can engage the user: "It’s no fun if it lags". A key driver of these advances is machine learning. Unlike earlier BCIs that relied on fixed models, machine learning allows systems to adapt and continuously learn from user signals and behavior, fostering a more intuitive human-machine interface through ongoing feedback and mutual learning.<br/> Despite these promising developments, significant challenges remain: real-time BCI systems demand high-performance hardware and software, require efficient wireless communication, robust processing capabilities, and adaptive algorithms. Furthermore, integrating these systems into real-world applications raises questions about usability and safety.<br/> This thesis addresses some of these challenges, particularly in the context of reducing cognitive load, addressing latency uncertainties and developing a modular BCI research framework. This work presents new insights and pathways toward functional BCI systems. Through experimentation and novel system architecture, this research lays a foundation for the future of BCI technology.}},
  author       = {{Tufvesson, Pex}},
  isbn         = {{978-91-8104-695-3}},
  language     = {{eng}},
  month        = {{11}},
  publisher    = {{Department of Automatic Control, Lund University}},
  school       = {{Lund University}},
  title        = {{Real-Time Brain-Computer Interfaces}},
  url          = {{https://lup.lub.lu.se/search/files/231478921/Pex_Tufvesson_-_HELA.pdf}},
  year         = {{2025}},
}