Publication Date
1-5-2026
Document Type
Article
Organizational Units
Daniel Felix Ritchie School of Engineering and Computer Science, Computer Science
Keywords
Brain maps, Objective-driven parcellation, Spatial brain analysis, Alzheimer's disease
Abstract
Brain parcellation schemes are fundamental to neuroimaging, yet general-purpose atlases may obscure the specific functional architecture relevant to a given cognitive task or clinical condition. This reflects a growing consensus that the “optimal” brain map is context-dependent. Here, we introduce a novel framework that validates this principle by generating task-optimized human brain parcellation maps directly from supervised learning objectives. Our method defines functional parcels by grouping brain regions based on the similarity of their contributions to a classifier's decision boundary for a specific goal (e.g., cognitive state decoding or clinical group separation). This approach prioritizes a region's discriminative role over simple signal homogeneity or spatial contiguity. We demonstrate that these objective-driven parcellations reveal a latent functional organization of the brain, an implicit task-relevant architecture defined not by signal homogeneity but by the shared discriminative role of brain regions. On Human Connectome Project data, our parcellations significantly improved cognitive state decoding, and on ADNI data, they enhanced Alzheimer's Disease classification. Beyond improving accuracy, the resulting parcellations exhibited unique neurobiological properties: they identified spatially coherent, high-resolution maps of task-relevant information that were obscured by standard atlases and showed a trade-off between task-specificity and overall signal homogeneity. These optimized maps generalized across independent datasets, highlighting that they capture robust principles of task-dependent brain organization. This work provides a framework for moving beyond universal atlases, enabling the generation of context-specific brain maps that offer a new window into the functional architecture underlying specific cognitive processes and disease states.
Copyright Date
1-5-2026
Copyright Statement / License for Reuse
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Rights Holder
Andrew Hannum and Mario A. Lopez
Provenance
Received from Elsevier
File Format
application/pdf
Language
English (eng)
Extent
20 pgs
File Size
4.88 MB
Publication Statement
Copyright is held by the Authors. User is responsible for all copyright compliance. This article was originally published as
Hannum, A., & Lopez, M. A. (2026). Task-Optimized Brain Parcellations Reveal Latent Functional Organization for Enhanced Connectivity-Based Neuroimaging Classification. NeuroImage, 325. https://doi.org/10.1016/j.neuroimage.2026.121689
Supplementary File Description
Graphical Abstract, JPG File, English(eng), 1 pg, 343 KB
Publication Title
NeuroImage
Volume
325
First Page
121689
ISSN
1053-8119
PubMed ID
41494385
Recommended Citation
Hannum, Andrew and Lopez, Mario A., "Task-Optimized Brain Parcellations Reveal Latent Functional Organization for Enhanced Connectivity-Based Neuroimaging Classification" (2026). Computer Science: Faculty Scholarship. 12.
https://digitalcommons.du.edu/computer_science_faculty/12
https://doi.org/10.1016/j.neuroimage.2026.121689
Included in
Analytical, Diagnostic and Therapeutic Techniques and Equipment Commons, Computer Sciences Commons, Neurosciences Commons