Page 196 - MEGIN Book Of Abstracts - 2023

the healthy control group. In the 30-80 Hz frequency mixed-norm methods) using computer simulations and
band, the strength of neuromagnetic source in patients auditory experiments. More importantly, we applied
with BECTS with FSIQ <80 was higher than that in the eFast-VESTAL to the presurgical evaluation of epilepsy.
FSIQ >80 group and the healthy control group. Our results demonstrated that eFast-VESTAL exhibited
a lower dipole localization error and/or a higher cor-
CONCLUSIONS The magnetic source inactivation of relation coefficient (CC) between the estimated source
the MFC and PCC regions during the interictal time may time series and ground truth under various conditions
be the reason for cognitive decline in early untreated of source waveforms. Experimentally, eFast-VESTAL
children with BECTS. Children with BECTS with cogni- displayed more focal activation maps and a higher CC
tive decline had a longer course of epilepsy and more between the raw and predicted sensor data in response
seizures. The magnetic source localization in the 4-8 Hz to auditory stimulation. Notably, eFast-VESTAL was the
frequency band may be a new imaging marker for the most accurate method for noninvasively detecting the
diagnosis of new BECTS. epileptic zones determined using more invasive stereo-
electroencephalography in the comparison.
Keywords: Accumulated source imaging, Benign child-
hood epilepsy with centrotemporal spikes, Cognitive IEEE transactions on bio-medical engineering (2021), Vol.
function, Magnetoencephalography, Multifrequency 68, No. 3 (32790623) (1 citation)
bands

Epilepsy & behavior: E&B (2020), Vol. 112 (32858366) (6 Towards the Automatic Localization of the Irritative
citations) Zone Through Magnetic Source Imaging (2020)

Luria, Gianvittorio; Duran, Dunja; Visani, Elisa; Rossi
Enhanced Fast-VESTAL for Sebastiano, Davide; Sorrentino, Alberto; Tassi, Laura;
Magnetoencephalography Source Imaging: From Granvillano, Alice; Franceschetti, Silvana; Panzica,
Theory to Clinical Application in Epilepsy (2021) Ferruccio

Zheng, Li; Sheng, Jingwei; Cen, Zhehang; Teng, Pengfei; Department of Mathematics, University of Genoa, Genoa,
Wang, Jing; Wang, Qian; Lee, Roland Robert; Luan, Italy. [email protected]; Department of Neurophysiology
Guoming; Huang, Mingxiong; Gao, Jia-Hong and Diagnostic Epileptology, IRCCS Foundation Carlo Besta
Neurological Institute, Milan, Italy; CNR - SPIN, Genoa, Italy;
ABSTRACT A novel magnetoencephalography source Epilepsy Surgery Center, Ospedale Niguarda, Milan, Italy
imaging approach called Fast Vector-based Spatio-
Temporal Analysis (Fast-VESTAL) has been success- ABSTRACT The present work aims at validating a
fully applied in creating source images from evoked Bayesian multi-dipole modeling algorithm (SESAME)
and resting-state data from both healthy subjects in the clinical scenario consisting of localizing the
and individuals with neurological and/or psychiatric generators of single interictal epileptiform discharges
disorders, but its reconstructed source images may from resting state magnetoencephalographic record-
show false-positive activations, especially under low ings. We use the results of Equivalent Current Dipole
signal-to-noise ratio conditions. Here, to effectively fitting, performed by an expert user, as a benchmark,
reduce false-positive artifacts, we introduced an and compare the results of SESAME with those of two
enhanced Fast-VESTAL (eFast-VESTAL) approach that widely used source localization methods, RAP-MUSIC
adopts generalized second-order cone programming. and wMNE. In addition, we investigate the relation be-
We compared the spatiotemporal characteristics of the tween post-surgical outcome and concordance of the
eFast-VESTAL approach to those of the popular distrib- surgical plan with the cerebral lobes singled out by the
uted source approaches (e.g., the minimum L2-norm/ methods. Unlike dipole fitting, the tested algorithms do

 ontents  Index 175
C