Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Differences in EEG Rhythm Energies Between PPA and Healthy Controls
Poster Session E, Sunday, September 14, 11:00 am - 12:30 pm, Field House
This poster is part of the Sandbox Series.
Kyrana Tsapkini1, Panteleimon Chriskos; 1Johns Hopkins Medicine, 2Laboratory of Medical Physics and Digital Innovation, , Faculty of Health Sciences, School of Medicine, Aristotle University of Thessaloniki, Thessaloniki, Greece
Introduction Primary progressive aphasia (PPA) is a rare neurodegenerative syndrome characterized by progressive deterioration of language abilities. It primarily affects language abilities including speech, comprehension and word-finding [1]. The current golden standard method for diagnosis of PPA is through clinical examination, Magnetic Resonance Imaging (MRI) or Positron Emission Tomography (PET) accompanied with lexical and/or language metrics [2]. In this work we aim to identify biomarkers that can be extracted through EEG recordings that allow the differentiation between PPA patients and healthy controls. Methods We recorded 3 minute, resting-state 32-channel EEGs from patients with PPA and compared them to healthy controls with eyes closed. The recorded EEG signal was preprocessed to remove noise. The steps involved were: 1. Mean value subtraction; 2. Digital filtering using five second order Butterworth filters (high-pass at 0.5 Hz, low-pass at 100 Hz, three bandstop filters at the 60 Hz power-line frequency and its first two harmonics); 3. Application of the Independent Component Analysis (ICA) method in order to remove components containing noise that could not be removed through filtering; and, 4. Epoching by segmenting the EEG recordings into 4 second epochs. After preprocessing, we conducted feature extraction by calculating the energy ratios of six EEG rhythms, specifically delta (0.5-4 Hz), theta (4-8 Hz), alpha (8-12 Hz), beta (12-20 Hz), gamma (20-50 Hz) and high-gamma (50-100 Hz). This was achieved through the Fast Fourier Transform. Two statistical analyses took place: 1. Each patient was individually compared with the control group, and, 2. Between the patient group and the control participants. Data were collected from 10 patients (5 logopenic variant PPA, 4 nonfluent variant PPA and 1 semantic variant PPA) (MoCA: 20.40±4.99) and 4 healthy controls at this preliminary stage (MoCA>29). In total there were 497 PPA epochs and 221 control epochs. We used 50 epochs per individual (or about 3.5 minutes) to minimize bias. Results We considered statistically significant differences between each patient (Welch’s t-test) compared to the control group those that had a p-value<0.05. Analysis 1: Individually for each patient, in decreasing patient numbers that achieved statistical significance the EEG rhythms are: theta (N=9, t=[-5.296, 3.957], p=[<0.001, 0.042]), alpha (N=8, t=[-3.25, 3.957], p=[<0.001, 0.017]), high-gamma (N=8, t=[-2.63, 3.955], p=[<0.001, 0.007]), beta (N=7, t=[-3.02, 3.036], p<0.001), gamma (N=7, t=[3.25, 3.957], p=[<0.001, 0.008]) and delta (N=6, t=[-5.296, 3.957], p=[<0.001, 0.01]). Analysis 2: For the comparisons between groups, 3 EEG rhythm energy ratios were found to be statistically significant: alpha (t=-4.496, p<0.001), gamma (t=5.664, p<0.001) and high-gamma (t=7.478, p<0.001). Conclusion These results show that patients with PPA present decreased alpha (as a group as well as in 7/8 patients) and increased gamma and high-gamma (as a group and in 7/8 patients) at the group level.Deficits in alpha rhythms have been associated with working memory deficits and deficits in high gamma have been associated with deficits in problem solving. In the next few months we intend to associate these associations with behavioral tasks performed by these participants, while increasing the patient and control sample size.
Topic Areas: Methods, Disorders: Acquired