MRI-based models of electric-field distribution for the modulation of primary progressive aphasia
Poster No:
45
Submission Type:
Abstract Submission
Authors:
Antoni VALERO-CABRE1, Xavier Corominas-Teruel2, Jeanne Salle3, Nicole Macias4, Souad Keichiri5, Maxime Janbon5, Michel Khachaturyan6, Clara Sanches7, Marc Teichmann8
Institutions:
1CNRS UMR 7225, Paris Brain Institute, Paris, Ile de France, 2Paris Brain Institute, Paris, Ile de France, 3COGMASTER program, University of Paris, Paris, Ile-de-France, 4Universitat Rovira i Virigili, Tarragona, Tarragona, 5Master BIP Université Sorbonne, Paris, Ile-de-France, 6Master iMind, Sorbonne Université, Paris, Ile-de-France, 7CNRS UMR 7225, Paris Brain Institute, Paris, Ile-de-France, 8Im2a, Hôpital de la Pitié-Salpetriêre, Paris, Ile-de-France
First Author:
Co-Author(s):
Introduction:
Transcranial direct current stimulation (tDCS) is a non-invasive technology used to modulate cortical activity in clinical settings [1]. Preliminary evidence suggests its outcomes are dramatically impacted by interindividual differences in head/brain structural features [2]. The optimization of tDCS parameters on the basis of personalized biophysical electric field (E-field) models could boost clinical efficacy. Unfortunately, under-performing tissue-segmentation algorithms limit their reliability, which remains controversial when applied to populations with cortical damage [3]. Additionally, the influence of specific tissue layers modified by pathological conditions, such as neurodegeneration, on tDCS current remains uncertain [4]. We here aimed to: (1) compare the accuracy of E-field distribution models based on automatic vs. manually MRI-segmentation approaches, (2) gauge the influence of head tissue layers on electrical current strength; and (3) assess their ability to predict cognitive modulation in patients with dementia.
Methods:
A cohort of n=16 patients diagnosed with semantic variant of primary progressive aphasia (sv-PPA) was stimulated with a single session of anodal tDCS (1.57mA, 0.06 mA/cm2, for 20 min) to the left Anterior Temporal Lobe (ATL). Guided with an MRI-based neuronavigation system, an anode was placed on a scalp site showing the shortest path to relevant ATL MNI coordinates [x=-53; y=4; z=-32] and a cathode over the right supraorbital region (AF8) [5]. language performance was assessed prior and following the tDCS session with a Semantic Association task [6]. A Finite Element Model (FEM) of the patient's head/brain tissue layers was built with the SimNIBS3.2.6 headreco pipeline [5]. In parallel, we manually segmented the layers of the model (White Matter, Gray Matter, CSF, Bone, Air, Eyes and Skin) and used this same tool to build a structural model for comparison. E-field simulations were conducted in both types of models and assessed with metrics assessing local and global E-field impact. Measures regarding the volume of the different tissue layers crossed by the E-Field were also estimated. Non-parametric statistics tested differences on tissue volumes and predicted E-field estimated on manually vs automatically segmented FEM models. Spearman correlations explored associations between structural measures, E-field values and changes in semantic abilities induced by anodal ATL tDCS.
Results:
We compared manually vs. automatically segmented head-models' E- total and normal (~tangential) E-field strength throughout the left temporal lobe or in a smaller ROI (10 mm radius sphere on ATL MNI coordinates) (Fig.1A&B). Significantly higher values in total E-field strength were found for manually, compared to automatically segmented FEM models (p=0.0131, Fig.2A). White matter (p=0.0062), CSF (p=0.0443) and skin (p=0.0003) volumes differed statistically between both types of models;) (Fig.2B). Moreover, cortical thickness (r=0.768; p=0.0374), CSF volume (r=-0.520; p=0.0386) for the whole temporal lobe significantly predicted total E-field strength, estimated with manual segmentation (Fig.2C). Unfortunately, no significant correlations between patient's semantic performance gains and E-field estimates were found for any of the two FEM structural models tested (Fig.2D).
Conclusions:
We conclude that MRI-based current distribution models built with automatic tissue segmentation algorithms remain suboptimal estimating E-field currents in atrophied brains. Additionally, we identified tissue layers (gray matter and CSF) modified by the pathology that impact the most E-field estimates predicted on the tDCS target. Unfortunately, E-field models failed to correlate with language performance gains. Our results highlight the need for individually customized stimulation strategies to achieve more efficient tDCS clinical interventions via MRI-based biophysically inspired computational models.
Brain Stimulation:
Non-invasive Electrical/tDCS/tACS/tRNS 1
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) 2
Keywords:
Aphasia
Cerebro Spinal Fluid (CSF)
Cognition
Neurological
Segmentation
Systems
Transcranial Magnetic Stimulation (TMS)
Treatment
Other - Transcranial Direct Current Stimulation
1|2Indicates the priority used for review
Provide references using author date format
[2] Liu A, Vöröslakos M, Kronberg G, Henin S, Krause MR, Huang Y, et al. immediate neuro-physiological effects of transcranial electrical stimulation. Nat Commun 2018;9.
[3] Unal G, Ficek B, Webster K, Shahabuddin S, Truong D, Hampstead B, et al. Impact of brain atrophy on tDCS and HD-tDCS current flow: a modeling study in three variants of primary progressive aphasia. Neurological Sciences 2020; 41:1781–9.
[4] Saturnino GB, Thielscher A, Madsen KH, Knösche TR, Weise K. A principled approach to conductivity uncertainty analysis in electric field calculations. Neuroimage 2019;188:821-34.
[5] Teichmann M, Lesoil C, Godard J, Vernet M, Bertrand A, Levy R, Dubois B, Lemoine L, Truong DQ, Bikson M, Kas A, Valero-Cabré A. Direct current stimulation of temporal poles boosts semantic processing in primary progressive aphasia. Annals of Neurology 2016, 80(5): 693-707
[6] Sanches C, Levy R, Benisty S, Volpe-Gillot L, Habert MO, Kas A, Ströer S, Dubois S, Kaglik A, Bourbon A, Migliaccio R, Valero-Cabré A, Teichmann A. Testing the therapeutic effects of transcranial direct current stimulation (tDCS) semantic dementia: A double blind, sham controlled, randomized clinical trial. Trials 2019 20(1):632.
[7] Nielsen JD, Madsen KH, Puonti O, Siebner HR, Bauer C, Madsen CG, et al. Automatic skull segmentation from MR images for realistic volume conductor models of the head: Assessment of the state-of-the-art. Neuroimage 2018;174:587–98.