Applications – Brain computer Interface and Deep Brain Stimulation
Key Words:
o Brain computer interface (BCI) = activates electronic or mechanical devices with
brain activity alone (Birbaumer, 2006).
o Locked in syndrome = having control of at least one muscle, followed by complete
locked in syndrome, having no control of any muscles.
o P300 = a stimulus driven EEG reflex response.
o Slow cortical potentials = slow event-related direct current shifts of the EEG.
o Deep brain stimulation (DBS) = electrical stimulation of the target brain regions to
help relieve symptoms of neurologic conditions.
o Stereotaxic surgery = holes drilled into the skull.
o Tics = sudden, rapid, recurrent, nonrhythmic motor movements or vocalisations,
generally preceded by urge.
o Plasticity = refers to the brains ability to reorganise neural pathways throughout the
life span, as a result of experience.
o Machine learning = collecting examples of EEG in calibration during which the user is
cued to perform repeatedly a small set of mental tasks. The data is used to adapt the
system to the specific brain signals of each user.
Key Studies:
Farewell & Donchin (1988): 4 healthy p’s, researchers called out names to them
followed by their own name, the brain then reacts to it which is the P300. It usually
takes 300ms for the brain to register its own name, a visible P300 can’t be detected
after a single exposure, will need to be repeated 20-30 times. Letter grids were
created to detect what letter the p was thinking of (non-invasive BCI, P300).
McCrane et al (2014): letter grids using 25 ALS p’s, there were over 35 target letters.
The L40 group had about 40% accuracy and the G70 group had 70% accuracy, the
differences were a lot to do with computer set-ups. Also, it was noted the L40 group
had a lot of visual impairments = presents a problem for visually induced P300 BCI’s
and audio tasks take a lot longer (non-invasive BCI, P300).
Birbaumer et al (1999): 2 ALS p’s using the operational learning approach for
physiological behaviour, perceived contingency = reinforced leading to reward (non-
invasive BCIs, slow cortical potentials).
Hammond (2011): lower frequency beta waves (12-15Hz) indicate the sensorimotor
rhythm, the sensorimotor rhythm reacts to real and imagined movement (non-
invasive BCI, sensory motor rhythms).
Wolpaw & McFarlane (2004): 4 p’s, 2 who are impaired had electrodes connected to
their scalps and asked to imagine moving their hands and feet, they needed training
to produce the correct signals, 6-12 weeks (non-invasive BCI, sensory motor
rhythms).
, Kennedy et al (2004): 2 ALS patients had glass microelectrodes filled with growth
factor implanted into the left motor cortex, measuring spike trains, after 8 tries one
p took under 5 seconds to achieve the desired outcome (invasive BCI).
Leuthardt et al (2004): 4 patients who had 32 subdural electrode arrays to identify
their focus of epileptic seizures. Each p performed 8 tasks (invasive BCI,
electrocorticogram).
Shrock et al (2014): most common DBS sites: pallidal (anteromedial globus pallidus
internus, posteroventral globus pallidus and globus pallidus externa) and
centromedian parafascicular complex.
Shrock et al (2014): review of DBS for Tourette’s: 30-80% improvement in tic
severity, 30-90% improvement in tic frequency but very small sample sizes and few
meet the criteria for FDA approval.
Heinrich et al (2004): contingent negative variation is an expectancy wave in the EEG,
most often recorded above the motor cortex. Contingent negative variation is low in
children with ADHD, training can improve this.
Banaschewski & Brandeis (2007): ADHD children trained in improving contingent
negative variation scored significantly lower of ADHD scores.
Kleutsch et al (2014): PTSD p’s had training for 30mins to lower their EEG alpha
below a threshold, lead to a significant pre to post increase in calmness and
significant difference in state anxiety. Using fMRI found increased pre to post
neurofeedback connectivity (subgenual anterior cingulate and middle frontal gyri
(neurofeedback to treat PTSD).
Reiter et al (2016): more studies are needed to show the efficacy of treating PTSD
with neurofeedback.
Panisch & Hai (2019): meta-analysis, most studies showed changes in at least 1
measure of PTSD symptomology. Current evidence is limited so research should aim
to: standardise approaches and outcome measures, control for comorbid conditions
and use randomised trials, but neurofeedback does show potential for a good
treatment for PTSD.
Birbaumer (2006): non-invasive EEG BCIs allow communication in paralysed and
locked in patients, but not in completely locked in patients, no firm conclusions
about the clinical utility of BCIs for the control of voluntary movement can be made
but invasive BCIs in healthy animals allowed them to reach and grasp. There are
newly developed fMRI and NIRS BCI’s which offer more opportunities.
Wichmann & DeLong (2016): the effects of DBS are dependent on targeting
sensorimotor portions of specific nodes of the basal ganglia-thalamocortical motor
circuit. Little evidence suggesting DBS in patients with movement disorders restores
normal basal ganglia functions, in stead it seems high frequency DBS replaces the
abnormal basal ganglia output with a tolerable pattern, helping to restore
functionality of networks.
Kuanqing (2016): in a lot of patients (who have major depression or bipolar disorder)
drug treatment doesn’t offer continuous effective symptom relief – DBS for
treatment resistant patients is being investigated especially due to recent safer
stereotaxic surgery and improving functional neuroimaging to map the affected
