Charles A. Ellis

The paper identifies altered default mode network connectivity dynamics in schizophrenia using a novel iterative feature-removal clustering approach on resting-state fMRI data.

This paper develops an active learning framework to personalize brain stimulation therapies, showing it outperforms random sampling in efficiency and accuracy across in silico and in vivo experiments.

This study identifies reproducibly important EEG markers of schizophrenia using an explainable multi-model deep learning approach and finds effects on beta, theta frequency bands, left hemisphere, and interhemispheric interactions.

This study identifies a subset of default mode network connectivity features in schizophrenia that reveal disorder-related dynamics not apparent when all features are included.

This study evaluates six EEG data augmentation methods for MDD diagnosis, finding that most do not improve performance beyond using duplicate data, but channel dropout augmentation does.

This study improves EEG-based MDD diagnosis using transfer learning by adapting representations from a sleep dataset to a higher sampling rate, increasing model accuracy and robustness.

This study introduces an explainable fuzzy clustering framework for analyzing dynamic functional network connectivity in schizophrenia, identifying 5 states and linking them to symptom severity.

This study introduces two novel convolutional neural networks for MDD diagnosis from EEG data, finding higher beta and delta power and right hemisphere correlation in MDD patients.

This study improves major depressive disorder diagnosis using raw multichannel EEG with transfer learning from single-channel sleep stage data, achieving better performance than previous methods.

This study develops a robust and explainable deep learning model for schizophrenia diagnosis using EEG data, showing improved performance and prioritization of specific brain regions and frequency bands.

This study combines explainable deep learning classification with clustering to analyze schizophrenia's effects on brain network dynamics, revealing reduced variability and temporally localized impacts.

This paper introduces an explainable fuzzy clustering method for analyzing dynamic functional network connectivity in fMRI data, showing it better identifies feature importance in schizophrenia compared to existing methods.

This study introduces a novel explainability approach for rs-EEG deep learning classifiers, revealing spatio-spectral interactions in MDD diagnosis.

This study identifies neuropsychiatric disorder subtypes and shows how a deep learning classifier's performance varies across these subtypes using dynamic functional network connectivity.

This paper introduces a framework for evaluating deep learning classifiers on raw EEG data, focusing on spatial, spectral, and temporal explainability, and applies it to MDD diagnosis.

This study develops novel methods to assess modality importance in multimodal electrophysiology classifiers, finding EEG is globally most important for sleep stages except stage 1, with subject-level differences influenced by sex, medication, and age.

This paper introduces a convolutional autoencoder-based clustering method for resting-state EEG data, identifying 8 EEG states in schizophrenia and linking time outside the dominant state to negative symptom severity.

This study integrates confidence estimation and explainability methods for schizophrenia diagnosis using rs-fMRI data, finding that MCBN is better suited for gradient-based explanations than MCD.

The paper examines the association between static and dynamic functional network connectivity and genetic risk of Alzheimer's disease using rs-fMRI and APOE data from cognitively normal individuals.

This study uses a deep learning classifier to subtype schizophrenia patients based on dynamic functional network connectivity, identifying 5 subtypes with distinct network interactions and symptom severity.

This study uses clustering and feature importance analysis on fMRI data to identify cognitive differences between brain network groups in a large population.

This study trains and compares random forest and SVM models to identify reproducible EEG biomarkers for schizophrenia using spectral features from resting state data.

This study applies explainable deep learning models to raw EEG data to diagnose schizophrenia, identifying key spectral and spatial features that align with prior research.

This study introduces a method to estimate uncertainty in gradient-based explanations for fMRI dFNC classification, finding significant differences in explanation relevance between schizophrenia and control groups.

This paper introduces a model visualization approach to analyze CNN filters for EEG waveforms and spectral features, demonstrating its use in sleep stage classification.

This study introduces a model visualization approach to explain CNNs' time and frequency features from raw EEG data, showing consistency with clinical sleep stage classification guidelines.

This paper introduces an activation maximization method for electrophysiology classifiers, showing it can identify spectral patterns in EEG sleep stages and reveal potential reliance on artifacts for awake classification.

This study classifies sleep stages using multimodal electrophysiology data and introduces an ablation method to assess modality importance, finding EEG is most important for most stages and EOG for REM and nonREM.

This paper introduces a local ablation method to explain multimodal classifiers, showing EEG's importance in sleep staging and identifying demographic effects on classification.