Classification and Extraction of Resting State Networks Using Healthy and Epilepsy fMRI Data

Abstract

Functional magnetic resonance imaging studies have significantly expanded the field's understanding of functional brain activity of healthy and patient populations. Resting state (rs-) fMRI, which does not require subjects to perform a task, eliminating confounds of task difficulty, allows examination of neural activity and offers valuable functional mapping information. The purpose of this work was to develop an automatic resting state network (RSN) labeling method which offers value in clinical workflow during rs-fMRI mapping by organizing and quickly labeling spatial maps into functional networks. Here independent component analysis (ICA) and machine learning were applied to rs-fMRI data with the goal of developing a method for the clinically oriented task of extracting and classifying spatial maps into auditory, visual, default-mode, sensorimotor, and executive control RSNs from 23 epilepsy patients (and for general comparison, separately for 30 healthy subjects). ICA revealed distinct and consistent functional network components across patients and healthy subjects. Network classification was successful, achieving 88% accuracy for epilepsy patients with a naïve Bayes algorithm (and 90% accuracy for healthy subjects with a perceptron). The method's utility to researchers and clinicians is the provided RSN spatial maps and their functional labeling which offer complementary functional information to clinicians' expert interpretation.

Resting-State Functional Connectome in Patients with Brain Tumors Before and After Surgical Resection

“…This can be achieved by several methods: an automatic or semiautomatic procedure with independent component analysis or through a spatial matching with respect to network templates 16,21 or manually with a seed-based approach, where predefined regions of interest (ROIs) are selected based on a-priori hypothesis using the Brodmann atlas coordinates. [1][2][3] Novel alternative methods, such as machine-learning approaches, 17,22 cortical parcellating approach, 8,23 or graph analyses, 7 are also used for easier-to-use methods in the clinical practice.…”

Epilepsy Focus Localization in Patients Utilizing BOLD Differences Related to Regional Metabolic Dynamics

“…The lack of cooperation of small children and patients with cognitive deficits to cope with task-related fMRI poses a pragmatic problem in the attempt to map their critical functions. For these reasons, the Independent Component Analysis (ICA) of the rs-fMRI has been utilized in different clinical centers to localize sensory-motor [7], visual [8], and language functions [9] [10]. One intermediate file obtained during the rs-fMRI processing with FSL MELODIC is the "mean" file; plainly, a signal average across the time points of the rs-fMRI sequence.…”

Pre-surgical Brain Mapping: To Rest or Not to Rest?

“…Multiple analysis techniques are available and, according to the method, different results can be obtained from the same dataset ( 11 , 24 , 25 ). Identification of the network is a critical step and can occur: (i) automatically or semi-automatically with Independent Component Analysis (ICA), a data-driven method commonly used: a network among a set of components is selected either visually ( 26 ) or through a spatial matching with respect to network templates ( 17 , 27 , 28 , 29 ); (ii) manually, with a seed-based approach, where pre-defined region-of-interest (ROI)s or seeds are selected based on a-priori hypothesis; (iii) with alternative methods, such as machine learning approaches ( 18 , 30 , 31 ), cortical parcellating approach ( 19 ) or graph analyses ( 15 ).…”