# The Physiological Effects of Transcranial Direct Current Stimulation ###### Will Strawson, Mara Cercignani --- # (Non-Invasive) Electrical Brain Stimulation #### Electro-Convulsive Therapy #### Transcranial Magnetic Stimulation #### Transcranial Direct Current Stimulation <aside class="notes"> * Neuromodulatory * Why tDCS? Cheaper, and easier to implement. These have been reciving lots of attention due to their ability to modulate cortical excitability. For example, theyve been foudn toeffect cogntivie processes as well as serve as likely candidates for depression treatment. </aside> --- # What is tDCS? --- <aside class="notes"> Involves passing a low, constant and direct current through a pair of electrodes (stimulating and reference) placed on the head. These are connected to a stimulator which delivers the current. This induces an intracerebral current flow which alters neuronal excitability, especially in the area under the stimulating electrode (here, the red electrode). The specific excitability change to occur depends on the type of stimulation: anodal stimulation, in which a positive current is passed through the circuit = increase in cortical excitability; In cathodal, negative = cortical inhibition. </aside> --- # Stimulate the depression away? #### Studies suggest tDCS is at least as effective as SSRIs, but with less side effects. ---- #### However, protocols not optimized - e.g. dose, strength etc (Yavari et al, 2017) ###### (Fregni et al, 2006b; Boggio et al, 2008; Loo et al, 2012; Knotkova et al, 2012; Brunoni et al, 2013; Brunoni et al, 2016) <aside class="notes"> * Cruailly - Protocals not optimised because the mechanisms behind the treatment are unknown </aside> --- ### Aim: To investigate the physiological changes induced by tDCS... #### ... may elucidate the underlying mechanisms. <aside class="notes"> * ... end goal - to help to imrpove treatment protocals </aside> --- # How? ### Three tDCS stimulation types [IV] ### 1) Sodium Imaging [DV1] * ##### Lack of previous research ### 2) Arterial Spin Labeling (ASL) [DV2] * ##### Anodal = Increase rCBF, Cathodal = Decrease in rCBF ###### (Lang et al, 2005; Zheng et al, 2011; Stagg et al, 2013; Sherwood et al, 2018) <aside class="notes"> * Why Sodium? tDCS alters resting membrane potential, relies on action potentials and thus Na-K channel * administration of carbamazepine, a sodium channel blocker, abolished short-duration aftereffects of anodal stimaltion (Liebetanz et al, 2002; Nitsche et al, 2003 * Why ASL? Confirm previous research + blood flow and neuronal excitation are well coupled </aside> ---- | | | | -------- | -------- | | - Acute Increases in rCBF | - Acute Increases in rCBF | | - **Net Decrease in rCBF**| . ###### Zheng et al (2011), 1.4 mA, stimulating electrode over precentral gyrus (C4) <aside class="notes"> Potentially a reason we found no difference is that there was still an increase in blood flow left over from cathodal stimulation... ...however, then should see a difference with sham? </aside> ---- # Hypotheses 1) **Anodal stimulation** will be associated with a **change in sodium concentration** 2) **Anodal stimulation** will be associated with an **increase in rCBF** 3) **Cathodal stimulation** will be associated with an **decrease in rCBF** <aside class="notes"> Undirectional NA hypothesis due to lack of previous research No change in Na due to cathodal simulation hypothesised becuase of carbamazepine studies </aside> --- # Methods ---- # Participants ### Six participants from UoS ---- # Design ##### **{** Scan (20 mins) ---> tDCS (20 mins) ---> Scan (20 mins) **} x3** ##### (1) Anodal, (2) Cathodal, (3) Sham <aside class="notes"> Order was counterbalenced between PS and single blinded 'Stimulation was done after scanning - not concurrent' </aside> --- # Procedure ### fMRI * ##### ASL (blood flow), Sodium, Structural (MPRAGE) ### tDCS * ##### Stimulating electrode on left M1, reference electrode on right OFC * ##### 2000 mA (anodal and cathodal) * ##### Sham: 15s up ramp, 30s 2000mA, 15s ramp down <aside class="notes"> For sham, 30s of 2000mA then nothing (? ask francesco ) **DONT GO DOWN, GO RIGHT!** </aside> ---- # Sodium Image Acquisition ##### **Sodium signal** directly proportional to **sodium concentration** ###### 4 phantoms: 30, 50, 75 and 120mM NaCl - ---- #### MATLAB script extracts sodium signal from each phantom, creating linear regression...  ## y = m x + c ### y = sodium concentration ### x = sodium signal <aside class="notes"> R squared = .99, good model! </aside> ---- #### Sodium is Difficult to Image! ###### Less sensitive + less abundant than hydrogen ###### Voxel size of 4mm³ ###### Madelin et al (2014) <aside class="notes"> Conventional MRI tuned to the resonant frequency of protons, however, can also be tuned to sodium nuclei Using siemens 3T, get 4mm3. Problems: Na has 4x less intrinsic sensitivity than hydrogen In lower abundance than hydrogen Biexponential relaxation behaviour = signal lost in few millisenconds Therefore, sdoium is thousands of times less sensitive Therefore, use very short echo times (we use cartesian FLORET sequence to sample K space) The NMR sensitivity of sodium is 9.27% of the proton sensitivity, which combined with low concentrations compared to water protons (110 M, for 2 protons/water molecule), leads to an average sodium NMR signal about 30,000 times lower than proton signal in brain. </aside> ---- # ASL Acquisition   ##### Protocal used: **pCASL** (Psuedo-Continuous ASL) ##### Train of very short RF-pulses to invert (label) water spins in a narrow band <aside class="notes"> To explain ASL, helps to look at PET... Uses magnetically labled water molecules as endogenous tracer Tagging performed proximal to the imaging volume 750 pulses each lasting 0.5ms over 1.5s Draws on the advanages of CASL and PASL - reliable perfusion images with high SNR Less taxing on hardware than CASL </aside> --- # Analysis ##### 1) Compute Δ images (for Na and ASL) within subjects *(Post - Pre)* ###### For each participant... Na ---> x3 ASL---> x3 <aside class="notes"> Top image - Na, Bottom image - ASL (MNI space) </aside> ---- ##### 2) Compare Δ images between condition across subjects *(Paired t-test)* ### Anodal vs Sham / Cathodal vs Sham / Anodal vs Cathodal <aside class="notes"> SPM Talk about partial volume correction </aside> --- # Results --- ## Sodium Analysis #### No significant differences ---- ## Anodal > Sham  ### p = 0.05 (FWE) ---- ## Anodal > Sham  ### No FWE ---- ## Anodal < Sham ###  ---- ## Cathodal > Sham ###  ---- ## Cathodal < Sham ###  ---- ## Anodal > Cathodal ###  ---- ## Anodal < Cathodal ###  --- # ASL Analysis ### No significant differences ---- ## Anodal > Sham  ---- ## Anodal < Sham  ---- ## Cathodal > Sham  ---- ## Cathodal < Sham  ---- ## Anodal > Cathodal  ---- ## Anodal < Cathodal  --- # Hypotheses 1) Anodal stimulation will be associated with a change in sodium concentration **[x]** 2) Anodal stimulation will be associated with an increase in rCBF **[x]** 3) Cathodal stimulation will be associated with an decrease in rCBF **[x]** --- # Discussion ---- ## The number of participants... #### n = 6 <aside class="notes"> Zheng et al (2011) used 6 PS in each condition however did VOI analysis - less harsh Other studies used at least double ours </aside> ---- ## The tDCS electrode placement... #### Tape measure used - not precise #### Electrode positioning may have differed between participants ###### However - Rich et al (2019); Nikolin (2018) <aside class="notes"> Potential downside of tDCS compared to TMS - can't locate specific motor cortices by use of MEPs </aside> ---- ## *Total* Sodium concentration measured... #### Can't distinguish between intra- and extra-cellular sodium concentrations #### If tDCS has an effect on the transmission of sodium between these two spaces, we wouldn't know! <aside class="notes"> People are trying to figure out how to do this however the technique is only feasable at 7T+ </aside> ---- ## The delay between tDCS and scan... ###### **{** Scan (20 mins) ---> tDCS (20 mins) *--(2-5mins)->* Scan (20 mins) **}**  #### Effect on rCBF and sodium concentration may have become diminished during this delay #### Previous studies use concurrent tDCS = No delay <aside class="notes"> At first didn't think delay would matter due to - Cortical excitability changes after tDCS (measured by TMS MEPs) found to be the same during and after stimulation (Nitsche et al, 2005) Zheng et al - in the scanner Stagg et al (2013) - found anodal stimulation lead to decrease in perfusion, therefore the delay matters We did try mitigate this issue by scanning Na first... </aside> --- # Future directions ---- ## More participants ---- ## Concurrent tDCS! --- ## Study suggests there are no physiological effects (for Na / rCBF) of tDCS... ## Type 2 error? --- ## What I've learnt... * tDCS * Sodium imaging * ASL imaging * SPM * MATLAB --- # Acknowledgements ###### Prof Mara Cercignani, Balazs Orzsik, Iris Asllani, Francesco Di Lorenzo, Jamie Ward. --- # Questions? <!-- .slide: data-background="#991c3f" --> --- # The lost slides ---- # Mood Disorders (e.g. depression) * #### Massive impact on society * #### Current Treatments Inadequate <aside class="notes"> * WHO(2017) - just under 1/20 people suffer from depression worldwide * SSRIs generally more effective than placebo however the sig increase risk of adverse events often outweighs the small benefits (Jacobsen, 2017) </aside> --- # We need something new! --- $$ x=10 $$