Brain–Computer Interface (BCI) is a technology that promises to directly connect the human brain with external devices by translating user intentions into control signals. The primary aim of this technology is to develop applications that enhance the quality of life for individuals suffering from various motor-related pathologies. To achieve this, several approaches have been explored, including the integration of human intention with robotic intelligence, resulting in shared control systems.\\ In this thesis, an existing Motor Imagery EEG-based BCI, successfully used to drive an electric wheelchair, has been integrated with a Hidden Markov Model (HMM) framework to merge BCI and environmental data, creating a more robust shared control system. Additionally, this framework introduces a "rest" class, expanding the classifier’s output from two to three classes. This approach was tested on three healthy subjects in a simulated environment (ROS-Gazebo) using three different configurations. Enhanced performance was observed when environmental data were incorporated into the process. Based on the promising results of this preliminary work, further refinement and implementation could lead to higher stability and improved performances.
Brain–Computer Interface (BCI) is a technology that promises to directly connect the human brain with external devices by translating user intentions into control signals. The primary aim of this technology is to develop applications that enhance the quality of life for individuals suffering from various motor-related pathologies. To achieve this, several approaches have been explored, including the integration of human intention with robotic intelligence, resulting in shared control systems.\\ In this thesis, an existing Motor Imagery EEG-based BCI, successfully used to drive an electric wheelchair, has been integrated with a Hidden Markov Model (HMM) framework to merge BCI and environmental data, creating a more robust shared control system. Additionally, this framework introduces a "rest" class, expanding the classifier’s output from two to three classes. This approach was tested on three healthy subjects in a simulated environment (ROS-Gazebo) using three different configurations. Enhanced performance was observed when environmental data were incorporated into the process. Based on the promising results of this preliminary work, further refinement and implementation could lead to higher stability and improved performances.
A novel Hidden Markov Model implementation for Brain-Computer Interface driven wheelchair
TONIOLO, SEBASTIANO
2023/2024
Abstract
Brain–Computer Interface (BCI) is a technology that promises to directly connect the human brain with external devices by translating user intentions into control signals. The primary aim of this technology is to develop applications that enhance the quality of life for individuals suffering from various motor-related pathologies. To achieve this, several approaches have been explored, including the integration of human intention with robotic intelligence, resulting in shared control systems.\\ In this thesis, an existing Motor Imagery EEG-based BCI, successfully used to drive an electric wheelchair, has been integrated with a Hidden Markov Model (HMM) framework to merge BCI and environmental data, creating a more robust shared control system. Additionally, this framework introduces a "rest" class, expanding the classifier’s output from two to three classes. This approach was tested on three healthy subjects in a simulated environment (ROS-Gazebo) using three different configurations. Enhanced performance was observed when environmental data were incorporated into the process. Based on the promising results of this preliminary work, further refinement and implementation could lead to higher stability and improved performances.File | Dimensione | Formato | |
---|---|---|---|
Toniolo_Sebastiano.pdf
accesso aperto
Dimensione
8.16 MB
Formato
Adobe PDF
|
8.16 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/80176