Summary: New brain-machine interface technology allows those who are immobile to control their wheelchairs through mind control. BMI allows users to traverse both normal and crowded environments after training.
source: Click on the cell
A mind-controlled wheelchair could help a paralyzed person gain new mobility by translating the users’ thoughts into mechanical commands.
On November 18 in the magazine iScienceresearchers show that quadriplegic users can operate mind-controlled wheelchairs in a crowded natural environment after training for a long time.
“We showed that mutual learning of both the user interface algorithm and the brain and machine are important for users to operate these wheelchairs successfully,” says Jose Del R. Millan, corresponding study author at the University of Texas at Austin. “Our research highlights a potential path to improving clinical translation of this non-invasive brain-machine interface technology.”
Milan and colleagues recruited three quadriplegics for a longitudinal study. Each of the participants underwent training sessions three times a week for a period of 2 to 5 months.
Participants wore a headgear that detects brain activities through electroencephalography (EEG), which will be converted into mechanical commands for wheelchairs via a brain-machine interface device.
Participants were asked to control the direction of the wheelchair by thinking about the movement of their body parts. Specifically, they needed to consider moving both hands to turn left and both feet to turn right.
In the first training session, three participants had similar levels of accuracy — when the device’s responses match the users’ thoughts — at about 43% to 55%. During the training period, the brain-machine interface device group experienced a significant improvement in accuracy in Participant 1, who had reached over 95% accuracy by the end of his training.
The team also saw Participant 3’s accuracy increase from 3 to 98% halfway through their training before the team updated their device with a new algorithm.
The improvement observed in participants 1 and 3 is related to the improvement in feature discrimination, which is the ability of the algorithm to distinguish the pattern of brain activity encoding for “going left” thoughts from that for “going right.”
The team found that the best discrimination feature resulted not only from the machine’s machine learning but also learning in the participants’ brains. The EEG of participants 1 and 3 showed clear shifts in brain wave patterns as they improved accuracy in the machine’s mind control.
“We see from the EEG results that the person has enhanced the skill to modulate different parts of their brains to generate a ‘left-oriented’ pattern and a different ‘right-oriented’ pattern,” Millan says. participants.”
Compared to participants 1 and 3, participant 2 did not have any significant changes in brain activity patterns throughout the training. Its accuracy increased only slightly during the first few sessions, which remained stable for the rest of the training period. Millan says he points out that machine learning alone is insufficient to successfully maneuver such a mind-controlled device
By the end of the training, all participants were asked to drive their wheelchairs through the crowded hospital room. They had to navigate obstacles such as a room divider and hospital beds, which were set up to simulate a real-world environment. Both Participants 1 and 3 completed the task while Participant 2 failed to complete it.
“It appears that for someone to gain such good control over the brain-machine interface that allows them to perform a relatively complex daily activity such as driving a wheelchair in a natural environment, it requires some neuroplastic reorganization in our cortex,” Millan says.
The study also emphasized the role of long-term training for users. Although Participant 1 performed exceptionally well in the end, he struggled in the first few training sessions as well, Millan says. The longitudinal study is one of the first to evaluate the clinical translation of a non-invasive brain-machine interface technique in quadriplegic subjects.
Next, the team wants to know why Participant 2 did not experience the learning effect. They hope to conduct a more detailed analysis of all participants’ brain signals to understand differences and possible interferences for people experiencing learning in the future.
About this search for neurotechnology news
author: press office
source: Click on the cell
Contact: Press Office – Cell Press
picture: The image is in the public domain
Original search: open access.
“Learning to control a BMI-driven wheelchair for severe quadriplegicsWritten by Jose del R. Millán et al. iScience
Learning to control a BMI-driven wheelchair for severe quadriplegics
- Three participants learned to drive a non-invasive BMI-powered wheelchair
- Direct transfer of acquired BMI skills to control a wheelchair
- Objective learning and machine intelligence are key factors in BMI compiled robots
Mind-controlled wheelchairs are an interesting assisted mobility solution applicable for total paralysis. Despite advances in brain-machine interface (BMI) technology, its translation remains elusive.
The primary objective of this study is to investigate the hypothesis that the acquisition of BMI skills by end users is central to the control of a brain-stimulated non-invasive intelligent wheelchair in real-world settings.
We demonstrate that three quadriplegic spinal cord injury users can be trained to operate a non-invasive, self-paced thought-controlled wheelchair and perform complex navigation tasks. However, users who showed increased decoding and feature discrimination performance, significant changes in neuroplasticity and improved BMI order latency achieved higher navigation performance.
In addition, we show that dexterous and sustained control of robots is possible through low degrees of freedom and discrete and uncertain control channels such as kinematic imagery and BMI, by blending human and artificial intelligence through co-control methodologies.
We hypothesize that subject learning and joint control are the two essential components that pave the way for a non-invasive translational BMI.