LEICI   25638
INSTITUTO DE INVESTIGACIONES EN ELECTRONICA, CONTROL Y PROCESAMIENTO DE SEÑALES
Unidad Ejecutora - UE
congresos y reuniones científicas
Título:
Modification of an electrical impedance model to study auditory processing in piglets after mTBI
Autor/es:
FERNANDEZ CORAZZA, MARIANO; OEUR, ANNA; MARGULIES, SUSAN; PALANISWAMY, MADURAN
Lugar:
San Diego
Reunión:
Conferencia; Biomedical Engineering Society 2020 annual meeting (BMES2020); 2020
Resumen:
Introduction: Mild traumatic brain injury (mTBI) has been known to affect auditory information processing [1]. Electrical impedance tomography (EIT) measurements have the potential to serve as a non-invasive diagnostic and prognostic tool for mTBI because it can detect neural processes on the millisecond (ms) scale. The aim of this study was to adapt an existing infant piglet brain model developed for studying hemorrhages into a source localization model of 4-week-old piglets. Such a model will enrich our understanding of the local amplitude and regional effects of electrical activity before and after mTBI. This work describes a proof of concept of a piglet model for source localization and presents exemplar data from a healthy and an injured animal.Materials and Methods: An infant piglet model, originally developed for studying intracranial hemorrhage [2], was scaled and modified to represent a 4-week old piglet brain with a 32-electrode array. Digitization of the electrode array on the piglet head was completed in ElectrodeWizzard using a series of images taken that captured 3-4 views of each electrode in the coronal, horizontal, and sagittal planes. The infant model was scaled to the 4-week old piglet dimensions. The 3-D electrode data were then co-registered to the model and the lead field matrix were modified to incorporate the 32-electrode array. The inverse solution was estimated using sLORETA algorithms. The EEG data applied to the model were taken from two 4-week old female piglets, one was healthy and the other was subject to a single sagittal head rotation using a HYGE injury device. Non-invasive EEG were collected using a 32-channel HydroCel electrode net (EGI, Philips Neuro). Each animal was subject to an oddball click train, comprised of 70 standard tones (800 Hz) and 30 target tones (1000 Hz). Data were analyzed using Netstation Tools (EGI, Philips Neuro) to filter (band pass: 0.1-30 Hz), segment (300 ms epochs; 50 ms before and 250 ms after stimulus), detect and replace bad channels (> 200uV). Baseline correction, artifact and eye blink removal via independent components analysis and waveform averaging were completed in EEGlab (V.14.12) and Matlab (V. R2018b). Waveforms were averaged such that each animal had a representative standard and target waveform. We hypothesize that the processing of the target tone would diminish after mTBI, therefore the target waveforms were simulated in the model to localize sources and images of brain activations at 50, 85 and 110 ms post-target stimulus (Figure 1).Results and Discussion:In Figure 1, the healthy animal (top panel) displayed an area of activation localized in the temporal region. This is an area consistent with auditory processing in piglets [3]. Decreased levels of activation were observed in the same area for the injured animal (bottom panel) across similar time points.Conclusions:This report demonstrates a proof of concept for source localization in a piglet model for use with EEG data capturing auditory oddball event related potentials. The exemplar data suggests that auditory processing is reduced after an mTBI and influences the region and magnitude of neural activities.