Brain map portal for alcohol use, obesity, temporal LOBE epilepsy, and alzheimer's
A high-performance computing system using Low-d M-ICA and CBMA generates low-dimensional virtual representations of brain networks, addressing the limitations of current neuroimaging methods to identify clinically relevant biomarkers for neurological and psychiatric disorders.
Patent Information
- Application Number
- PCT/US2025/061242
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-12-23
- Filing Date
- 2025-12-23
- Publication Date
- 2026-07-02
AI Technical Summary
Current neuroimaging methods struggle to identify robust, context-specific neural networks governing individual behaviors or diseases due to limitations in generalizability, computational expense, and lack of theoretical basis for low-dimensional applications, making it difficult to develop clinically relevant neuroimaging biomarkers.
A high-performance computing system employing a Low-dimensionality Meta-analytic Independent Component Analysis (Low-d M-ICA) algorithm analyzes neuroimaging data to discover and validate network-based brain models, using multivariate coordinate-based meta-analysis (CBMA) to generate low-dimensional virtual representations of brain networks applicable to individual patients.
Enables the discovery and validation of clinically relevant neuroimaging biomarkers for various disorders, including epilepsy and psychiatric conditions, by accurately delineating context-specific neural networks, overcoming computational and generalizability challenges.