Y P O C Computational Modeling to T O Understand tDCS and tACS N O D Flavio Frohlich E University of North Carolina - Chapel Hill S Department of Psychiatry Department of Cell Biology and Physiology A Department of Biomedical Engineering E Department of Neurology Neuroscience Center L P www.facebook.com/FrohlichLabUNC
Y P COI UNC has a patent on (feedback) (t)ACS for modulation of cortical O oscillations as a therapeutic C intervention. FF is lead inventor (US PR App 61/899,954). UNC has determined the absence of a conflict T of interest (COI) for the work O presented here and has determined a “COI with administrative N considerations” for the clinical trials in the Frohlich Lab due to the use of hardware designed in the Frohlich O Lab. D • We are working on our own E device hardware. • I am writing a textbook S “Network Neuroscience” for Elsevier. A • I speak with people from E Neosync and Tal Medical and have received travel support. L • We use Neuroconn devices. P • My preferred brain stimulation modality is espresso.
Y P O C If your tDCS/tACS study only T uses behavioral outcomes, O N either (1) you hit the jackpot and your original hypothesis got O confirmed D E or (2) your results disagree with S your hypothesis, so you ??? A Sellers et al. 2015 E L P
Y P VERTICAL INTEGRATION O C T Patients Clinical Trials O N Brain Stimulation, COMPLEXITY Human Neurophysiology O D In vivo (Animal) E Electrophysiology S A In vitro (Animal) Electrophysiology TRACTABILITY E L Model Systems P Computer Simulations
Y P TRANSCRANIAL CURRENT STIMULATION O STUDY DESIGN C T O Behavioral N Network Target Target Target Engagement O D E S A E L P
Y P TARGET ENGAGEMENT O C T How do we best engage a O network target? N O D We need to understand what E the effect of stimulation is on S the brain in terms of A neurophysiology . E L P
Y P OUTLINE O C 1. Cellular Effects T O N 2. Spatial Targeting O D 3. Targeting Network Dynamics E S A E L P
Y P ELECTRIC FIELDS O C T O N How do electric fields change O electric signaling in neurons? D E S A E L P
Y P O “Anodal” “Cathodal” Depolarized Soma Hyperpolarized Soma C Hyperpolarized Dendrite Depolarized Dendrite T O N O D E S A E L P
Y P CABLE EQUATION O C T O N O D E S A E L P Frohlich and McCormick. 2010
Y P NEURONAL MORPHOLOGY AND STATE O C Change in somatic membrane voltage: T O • Increases with cable length. • Decreases with membrane conductance. N • Increases with cable diameter. O A B D E vs. S A E L P Radmann et al. 2009
Y P O C T O N O D Change in somatic membrane E voltage can be modeled as a sub- S threshold somatic current injection. A E L P Frohlich and McCormick. 2010
Y P SUMMARY CELLULAR EFFECTS O C T Weak electric fields change the membrane O voltage of neurons. The effect on an individual neuron depends on the field, the N neuron, and the spatial relationship O between the two. D To study network-level effects, an adjusted E somatic current injection can be used in S computational models. A E L P
Y P SPATIAL TARGETING O C T The electric field caused by current application • O is a function of the electric conductivity of the Resistivity tissue. N Tissue [Ohm cm] Mathematically, the so-called Laplace equation • Copper 2e-6 is numerically solved to determine the electric O potential from the current application. CSF 64 D The current application is modeled as a • boundary condition. Cortex 350 E The key parameter is the conductivity that • White Matter 650 greatly differs between tissues. S Bone 8,000-16,000 A Current flow the strongest in skin and • cerebrospinal fluid (shunting). E L P
Y P IMPLEMENTATION O C • MR Scan T • Tissue segmentation O • Numerical solution based on dividing head in a N large number of small compartments (e.g. finite elements). O D 1. Develop you own code, typically using MR E analysis tools and a physics simulator. 2. Collaborate with groups that developed such a S tool. A 3. Buy tool. E L P
Y P O C T O N O D E S A E L P
Y P O C T O N O D E S A E L P
Y P O C T O N O D E S A E L P Modeling performed by Angel Peterchev Sellers et al 2015
Y P O C T O N O D E S A E L P
Y P SUMMARY: SPATIAL TARGETING O C T • MR scan + Segmentation + EF O modeling = Spatial Targeting N • Conventional electrodes (scale: cm) O cause relatively broad electric field D distributions. E • Electric fields are not only superficial. S • Smaller (and more) electrodes may A provide better spatial targeting. E L P
Y P O STRUCTURE DYNAMICS C T O N O D E S A E L P BEHAVIOR
Y P MODELING DYNAMICS O C T O N O D E S A E L P Frohlich 2014
Y P OSCILLATIONS O C T O N O D E S A E Caution: Most tACS literature refers to the L peak-to-peak amplitude as amplitude . P
Y P NETWORK DYNAMICS O C Raw Trace Spectrum T O 1. Raw trace. N 2. Spectrum: Power as a function of frequency. O 3. Spectrogram: Spectrum as D a function of time. 4. Coherence: Interaction E between two sites as a S function of frequency. A E L P
Y P O C 1. Raw trace. T 2. Spectrum: Power as a function of frequency. O 3. Spectrogram: Spectrum as a function of time. N O D Raw Trace E Spectrogram S A E L P
Y P 1. Raw trace. O 2. Spectrum: Power as a function of frequency. C 3. Spectrogram: Spectrum as a function of time. 4. Coherence: Interaction between two sites as a function T of frequency. O N O D E S A E L P
Y P TARGETING BRAIN NETWORK DYNAMICS O C T Berger 1929 O N O Neuroconn Write / Input Read / Output D tACS EEG E S A E L P Transcranial Alternating Current Stimulation (tACS)
Y P NATURALISTIC ELECTRIC FIELDS O C T O N O D E S A E L P Frohlich and McCormick. 2010
Y P PHASE SYNCHRONIZATION O C T O N O D E S A E Detuning: Difference between natural (endogenous) and L stimulation (external) oscillation frequency. P
Y P PHASE SYNCHRONIZATION O C T O N O D E S A E L P
Y P ARNOLD TONGUE O C T O N O D E S A E L P Frohlich 2014
Y P SPIKING NEURAL MODEL (NETWORK) O C T O N O D E S A E L P Ali et al. 2013
Y P SPATIO-TEMPORAL DYNAMICS O C T O N O D E S A E L P Ali et al. 2013
Y P O C T O N O D E S A E L P Ali et al. 2013
Y P STIMULATION PHASE O C T O N O D E S A E L P Ali et al. 2013
Y P HOTSPOTS O C T O N O D E S A E L P Ali et al. 2013
Y P NETWORK-LEVEL MECHANISM O C T O N O D E S A E L P Ali et al. 2013
Y P CELLULAR-LEVEL MECHANISM O C T O N O D E S A E L P Ali et al. 2013
Y P O TARGETING A C SUBPOPULATION T O N O D E S A E L P Ali et al. 2013
Y P NETWORK RESONANCE O C T O N O D E S A E L P Ali et al. 2013
Y P PHASE SLIPPING O C T O N O D E S A E L P Ali et al. 2013
Y P INTERACTING NETWORKS O C T O N O D E S A E L P Kutchko and Frohlich 2013
Y P MULTISTABILITY O C T O “Rapid Fire” “Slow Propagating” “Spiral Waves” N O D E S A E L P Kutchko and Frohlich 2013
Y P STATE SWITCHING BY tACS O C T O N O D E S A E Kutchko and Frohlich 2013 L P
Y P TARGET: ALPHA OSCILLATIONS O C T O N O D E “Offline” state, long-range • S functional connectivity, gating. A Neurofeedback, rTMS (10 Hz), tACS E • L of visual cortex to modulate P perception, Neosync, etc.)
Y P THALAMUS: ALPHA, GAMMA, SPINDLES O C Awake (“online”) Awake (“offline”) NON-REM sleep T Gamma Oscillations Alpha Oscillations Spindles O N O D E S A E L P
Y P COGNITIVE ENHANCEMENT O C T “increased alpha power during creative O ideation is among the most consistent findings in neuroscientific research on N creativity” (Fink and Benedek, 2010) O High Creative Ideation Low Creative Ideation D E S A E L P Lustenberger et al. (2015)
Y P ENHANCING CREATIVITY O C T O N O D E S A Blinding was successful (p > 0.2). • E 10 Hz tACS significantly enhances creativity as measured by the Torrance • L Test of Creative Thinking (7.45 % ± 3.11 % S.E.M.; F 1,16 = 5.14, p = 0.036). P No enhancement with 40Hz-tACS.. • Lustenberger et al. (2015)
Y P STIMULATION ARTIFACT SUPPRESSION O C T Signal contaminated by Cleaned signal after stimulation artifacts artifact suppression O N O D E Artifact S A E Spectra showing peaks Spectra of cleaned signal L corresponding to artifacts shows elimination of peaks P
Y P OSCILLATION ENHANCEMENT O C T O N O D E S A E L P
Y P STATE-DEPENDENT MODULATION O C T O N O D E S A E L “Eyes Closed” “Eyes Open” P
Recommend
More recommend