Knowledge Vault 3/15 - G.TEC BCI & Neurotechnology Spring School 2024 - Day 2
The endovascular neural interface: from proof of concept to the clinical trial
Sam John, University of Melbourne (AU) & Hackathon Host
<Resume Image >

Concept Graph & Resume using Claude 3 Opus | Chat GPT4 | Llama 3:

Sam John
Paralysis impacts 50 million
globally. 1
Brain-machine interfaces aid paralyzed
individuals. 2
Sensors: invasive to less
invasive. 3
Stent electrodes: less invasive
recording. 4
Intravascular recording concept
since 1973. 5
Stentrode: chronically implantable
recording device. 6
Catheter deploys stentrode in
vessels. 7
Sheep studies: long-term
recording, signal improvement. 8
Stentrodes allow tissue growth,
remodeling. 9
Blood flow: initial expansion,
then reduction. 10
Modeling: strut thickness, growth,
shear stress. 11
Preclinical data enabled human
trials. 12
Stentrode: wireless control at
home. 13
Concept to human trial:
iterative journey. 14
Participant controls computer via
stentrode. 15
Participant uses stentrode for
communication, entertainment. 16
Collaborators involved in decade-
long development. 17
Early prototypes: commercial
stents, external electrodes. 18
Current stentrodes: thin-film multi-
metal deposition. 19
Minimum vessel diameter:
4-5 mm. 20
Human trials: superior sagittal
sinus implantation. 21
ALS recipients: potential
lifelong implantation. 22
Subdural arrays: higher
bandwidth than others. 23
Decoding performance: comparable
in sheep. 24
ECG artifact issue addressed
through referencing. 25
No major adverse events,
minor discomfort. 26
Future research: neuromodulation,
brain stimulation. 27
Arterial implantation: additional
risks, safety studies. 28
Challenges: decoding complexity,
long-term safety. 30

Resume:

1.-Paralysis affects about 50 million people globally due to various conditions like stroke, spinal cord injury, ALS, and multiple sclerosis.

2.-Brain-machine interfaces can help people with severe paralysis, consisting of a sensor, decoder, and assistive technology.

3.-Sensors range from invasive intracortical electrodes recording single neurons to less invasive epidural electrodes recording from many neurons.

4.-Stent electrodes placed in brain blood vessels without craniotomy could record neural signals while being less invasive than intracortical electrodes.

5.-The idea of intravascular neural recording has been around since 1973, with demonstrations in 1998, but no chronic implantable device existed.

6.-The stentrode is a stent with electrodes that can be implanted through blood vessels and remain chronically to record neural signals.

7.-A catheter is used to deploy the stentrode in the desired brain blood vessel location.

8.-Sheep studies showed stentrodes could record somatosensory evoked potentials over 1 year, with signals improving as the device incorporated into the vessel.

9.-Histology and imaging revealed stentrodes allowed blood vessel remodeling and tissue growth similar to foreign body response to conventional brain electrodes.

10.-Blood flow measurements showed an initial expansion of the vessel diameter post-implantation followed by a reduction to slightly below baseline by day 28.

11.-Computational modeling revealed a relationship between stentrode strut thickness, tissue growth, and wall shear stress that warrants further optimization.

12.-Preclical data enabled the first human stentrode trial, with the device delivered through the jugular vein and placed in the superior sagittal sinus.

13.-The stentrode connects to a wireless transmitter in the chest, allowing motor intention decoding for computer cursor control at home.

14.-The journey from initial concept to first human trial spanned many years and iterative improvements.

15.-A participant is shown controlling a computer using an eye tracker for cursor movement and the stentrode for "click" commands.

16.-Another participant uses the stentrode alone for cursor control and clicking to use communication and entertainment applications.

17.-Many collaborators were involved in the stentrode development over the past decade, with trials ongoing in the US.

18.-Early stentrode prototypes used commercially available stents with externally attached platinum electrodes and wires.

19.-Current stentrodes use monolithic thin-film multi-metal deposition, with polymer stents under investigation but not yet tested in humans.

20.-Minimum blood vessel diameter for stentrode implantation is estimated around 4-5 mm, with research needed on use in smaller vessels.

21.-Human stentrode trials so far have only implanted in the superior sagittal sinus, but various cerebral veins were tested in animals.

22.-Stentrode recipients with ALS have been implanted over 1 year, with the potential for lifelong implantation based on intracranial stenting for other conditions.

23.-Subdural arrays showed higher signal bandwidth than epidural and stentrode arrays, attributable to closer proximity to the brain.

24.-Decoding performance of epidural, subdural and stentrode arrays were comparable in sheep, but expected to be lower than intracortical arrays.

25.-ECG artifact was an issue with early stentrode recordings but addressed through improved referencing in later trials.

26.-No major adverse events related to stentrode implantation were observed, only minor discomfort at the implantation site in some cases.

27.-Neuromodulation and brain stimulation are key areas of future research for stentrodes, which could enable less invasive treatment of conditions like Parkinson's disease.

28.-Arterial implantation of stentrodes carries additional risks compared to venous implantation and requires long-term preclinical safety studies.

29.-Tissue overgrowth has not impacted stentrode chronic recording, with signals tending to improve over the first 14-28 days post-implantation.

30.-Key challenges remain in decoding more complex signals from stentrodes and understanding the long-term safety of intravascular neuromodulation.

Knowledge Vault built byDavid Vivancos 2024