BrainStorm: Highly Extensible Software for Advanced Electrophysiology and MEG/EEG Imaging

点击翻译按钮获取中文摘要

PROJECT SUMMARY We are requesting a supplement to R01 EB026299 (Brainstorm: Highly Extensible Software for Advanced Electrophysiology and MEG/EEG Imaging) under NOT-AG-22-025 to extend our Brainstorm software to develop an open-source Alzheimer's Disease (AD) analytical toolkit (AD Toolbox). The new software features will be designed to promote the use of multimodal electrophysiology (in conjunction with other imaging and data collection methods) in staging, understanding, and treating Alzheimer's Disease. R01 EB026299 provides support for developing Brainstorm, a Matlab/Java multi-platform (Linux, macOS, Windows) software environment for analysis and visualization of electrophysiological (e-phys) data. This open-source software plays an increasingly enabling role in clinical and cognitive neuroscience research as reflected in the following: 2,300 published articles reporting analyses using Brainstorm, >36,000 user accounts, 18,000 downloads/month, 45,000 posts on the online forum, and 2,500 researchers who have attended in-person or online Brainstorm training events. The research proposed here is fully within the scope of the current grant, as summarized in the Specific Aims of the competing renewal for R01 EB026299: “Under this renewal, we will continue to provide support, documentation, and training to Brainstorm users while exploring and developing innovative methodologies, computational tools, and integrated software-solutions relevant to current analysis and research involving e-phys data.” The neuropathology of AD is characterized by staged spreading of neuronal cell loss, neurofibrillary tangles, and plaques in multiple cortical and subcortical brain regions. The build-up of tau protein has now become another hallmark of AD pathophysiology. The development of PET ligands for both tau and - amyloid has resulted in increasingly sensitive tests for early-stage and preclinical AD. However, PET studies are expensive and do not shed direct light on the impact of the build-up of these proteins, or AD in general, on neuronal function. We will leverage current and planned computational tools within Brainstorm to develop methodologies for the use of EEG/MEG data in the early detection of AD and in understanding its neurophysiological impact on behavior and cognition as the disease progresses. We will use the longitudinal PREVENT-AD database (which includes MRI, -amyloid and tau PET, and MEG studies, plus medical records and extensive cognitive testing) to develop and test methodology and software. The AD Toolbox developed under this supplement will include the following two major components: (i) Computational methods for data processing, statistical analysis, and machine learning (ML) at the whole-brain level that will allow users to extract spatio-temporal features from MEG and EEG data and study their relationship with spatial protein maps of beta- amyloid and tau (PET), as well as relevant structural measures, such as cortical thickness and regional brain volume (MRI); (ii) Computational methods for localized analysis of electrophysiology. We will develop tools for investigating electrophysiologic markers, focusing on the entorhinal and perirhinal cortex and hippocampus .