The End Of Knowledge - Vault 3 - G.TEC BCI & Neurotechnology Spring School 2024 - Day 2

graph LR classDef bonnet fill:#f9d4d4, font-weight:bold, font-size:14px; classDef objectives fill:#d4f9d4, font-weight:bold, font-size:14px; classDef stimulation fill:#d4d4f9, font-weight:bold, font-size:14px; classDef potentials fill:#f9f9d4, font-weight:bold, font-size:14px; classDef challenges fill:#f9d4f9, font-weight:bold, font-size:14px; A[Francois Bonnetblanc] --> B[ cortical responses guide
tumor surgery, electrogenesis. 1] A --> C[Objectives: intraoperative connectivity, monitoring,
evoked electrogenesis, neuromodulation. 2] A --> D[Macro-stimulation: less invasive, selectivity issues.
Early research combined techniques. 3] A --> E[DCR in neurosurgery. CCP maps
epileptic connectivity. Tumor use. 4] A --> F[Surgical issues: functional margins,
avoiding disconnection, online monitoring. 5] F --> G[Brain-to-brain potentials expand monitoring,
overcome imaging limits. 6] A --> H[Intraoperative challenges: many electrodes,
artifacts, environment. Careful technique. 7] A --> I[Stimulation: bipolar, 1-10 Hz, 0.25-3 mA.
High sampling, minimal filtering. 8] A --> J[Key potentials: DCR, ACEP, CCP. 9] J --> K[Literature: N1 consistent across
DCR, ACEP, CCP. P0, N2 sometimes. 10] J --> L[Despite stimulation differences,
N1 reflects cortical output. 11] J --> M[Long-range: delay, attenuation, dilation
vs short-range. Conduction velocity. 12] J --> N[Non-canonical waveforms: polarity inversion,
oscillations. Not well understood. 13] J --> O[N1: summation of excitatory
postsynaptic potentials. Linear with unitary. 14] O --> P[N1 similar for DCR, ACEP.
Some EPSP differences. 15] O --> Q[DCR gamma activity during N1
relaxation vs ACEP. Horizontal connections. 16] J --> R[P0: summation of synchronized
action potentials. Direct connectivity. 17] R --> S[P0 calculates conduction velocities.
Difficult to measure. 18] S --> T[Velocities from P0 realistic.
N1 underestimates velocity. 19] A --> U[Modeling axonal activation: volume,
properties, orientation. Being researched. 20] A --> V[Challenges: resolution, activation 'hotspot',
artifacts. Multiple measures validate. 21] A --> W[Good recordings: high sampling,
amplitude, no filters. Polarity alternation. 22] A --> X[Classical parameters not well
characterized. Effect on response. 23] X --> Y[High frequency: negative drift,
N1 attenuation. Filters obscure. 24] A --> Z[Future: data fusion, real-time
connectivity, resection margins. 25] Z --> AA[Underlying electrophysiology needs understanding.
Source localization, generator modeling lacking. 26] A --> AB[Team: students Clotilde, Felix, Olivier.
Clinicians Montpellier, Paris, Japan. 27] A --> AC[Surgery: 0.2-3 mA bipolar.
Lower than epilepsy grids. 28] AC --> AD[No routine impedance. Artifact
prevents pulse recording. Low frequencies. 29] A --> AE[Focus on electrophysiology. Future:
pain, functional networks. 30] class A,B bonnet; class C objectives; class D,H,I,W,X,Y,AC,AD stimulation; class E,F,G,J,K,L,M,N,O,P,Q,R,S,T potentials; class U,V,Z,AA,AE challenges;

Resume:

1.-Francois Bonnetblanc discussed using direct cortical responses and cortical evoked potentials to guide brain tumor surgery and understand evoked electrogenesis.

2.-Key objectives are identifying anatomical connectivity intraoperatively, reinforcing intraoperative neural monitoring, and understanding evoked electrogenesis for effective brain neuromodulation.

3.-Macro-stimulation is less invasive than micro-stimulation but has selectivity issues. Early research combined surface recordings with invasive unitary extracellular recordings.

4.-Goddring applied the DCR technique to neurosurgery. Matsumoto uses CCP to map connectivity in epileptic patients. Current use is for tumor surgery.

5.-Surgical issues include taking margins in functional tissue and avoiding disconnection syndromes. Online real-time functional monitoring is used in awake surgeries.

6.-Brain-to-brain evoked potentials are being developed to expand intraoperative neural monitoring, especially under general anesthesia since imaging guidance has limitations.

7.-Measuring good evoked responses intraoperatively is complex due to many electrodes, stimulation artifacts, and the surgical environment. Careful technique is required.

8.-Stimulation uses a bipolar electrode at 1-10 Hz and 0.25-3 mA. High sampling rates (19.2 kHz) are used with minimal filtering.

9.-Key evoked potentials are DCR (stimulation and recording on same gyrus), ACEP (white matter stimulation, cortical recording), and CCP (cortico-cortical).

10.-Literature shows some variation in waveforms but a consistent N1 negative potential across DCR, ACEP and CCP. P0 and N2 also sometimes seen.

11.-Despite stimulation differences, canonical waveforms with N1 are seen for DCR, ACEP and CCP, mainly reflecting the cortical output.

12.-For true long-range potentials, a delay, attenuation and dilatation should be seen compared to the short-range potential due to conduction velocity.

13.-Some non-canonical waveforms are occasionally seen, like polarity inversion or oscillations, but are not well understood. Most are the canonical N1 waveform.

14.-The N1 is considered a summation of excitatory postsynaptic potentials at apical dendrites. Linear relationship exists with unitary extracellular responses.

15.-Despite stimulation differences, the N1 waveform is similar for DCR and ACEP, reflecting activation of large diameter axons. Some EPSP differences seen.

16.-Time-frequency analysis shows increased DCR gamma activity during N1 relaxation phase compared to ACEP, possibly reflecting activation of horizontal cortical connections.

17.-The P0 potential, occurring before 8 ms, is considered a summation of synchronized action potentials on directly activated large axons.

18.-The P0 is a marker of direct anatomical connectivity and can be used to calculate conduction velocities, but is difficult to measure.

19.-Conduction velocities calculated from the P0 are realistic based on axon diameters. The N1 peak should not be used as it underestimates velocity.

20.-Modeling axonal activation from stimulation is being researched. Factors include the conductive volume, axon properties, and stimulation orientation relative to axons.

21.-Challenges include limited resolution, finding the "hotspot" of activation, and removing stimulation artifacts. Using multiple measures can help validate long-range potentials.

22.-High sampling rates, amplitude resolution, no hardware filters are important for good recordings. Alternating stimulation polarity may be helpful but under-researched.

23.-Classical stimulation parameters like intensity, pulse width, and frequency are not well characterized in terms of effect on the evoked response.

24.-Increasing stimulation frequency causes a slow negative drift and attenuation of the N1 response. High pass filters can obscure this effect.

25.-Future research directions include data fusion with imaging, measuring connectivity in real-time, and using evoked responses to guide tumor resection margins.

26.-However, the underlying electrophysiology needs to be better understood first. Source localization and generator modeling are lacking compared to spontaneous EEG research.

27.-The research team includes PhD students Clotilde, FÃlix and Olivier, collaborating with clinicians in Montpellier, Paris and Japan (Prof. Matsumoto's group).

28.-Stimulation intensities are 0.2-3 mA with a bipolar probe for tumor surgery, lower than epilepsy grids due to current density considerations.

29.-Impedance is not routinely measured. Artifact prevents recording during the stimulation pulse itself. Frequencies are kept low to avoid after-discharges and seizures.

30.-The team uses standard ECoG and is focused on the electrophysiology. Future work may look at pain and other functional networks.

Knowledge Vault built byDavid Vivancos 2024