AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

A New Brain Network Construction Paradigm for Brain Disorder via Diffusion-Based Graph Contrastive Learning

Yongcheng Zong, Qiankun Zuo, Michael K. Ng, Baiying Lei, Shuqiang Wang

TPAMI 2024 · Graph learning › Graph self-supervised learning · approach: Contrastive learning with negatives and views · 89 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add BNOM graph contrastive optimization (ablation)BRAM pipeline · ADNI 4-classAUC: 0.8914 -> 0.9781Improved
add BRAM region-aware module (ablation)BNOM pipeline · ADNI 4-classAUC: 0.9536 -> 0.9781Improved
add BRAM region-aware module (ablation)BNOM pipeline · ADNI 4-classACC: 0.9138 -> 0.9327Improved
use diffusion-based BRAM (mu=0.8) instead of ANTs for registrationregistration · ADNIDice: 0.694 -> 0.798Improved
raise smoothness weight mu from 0.8 to 1.0 (ablation)BRAM · ADNIDice: 0.798 -> 0.786Hurt
lower smoothness weight mu from 0.8 to 0.3 (ablation)BRAM · ADNIDice: 0.798 -> 0.671Hurt
raise reference network coefficient alpha from 0.1 to 0.7 (ablation)DGCL · ADNI AD vs NCACC: 0.953 -> 0.971Improved
raise alpha from 0.7 to 0.9 (ablation)DGCL · ADNI AD vs NCACC: 0.971 -> 0.965No clear effect
use DGCL instead of PANDA for brain network constructionbrain network construction · ADNIreliability (ICC): 0.73 -> 0.79Improved
use DGCL instead of PANDA for brain network constructionbrain network construction · ADNItime (s): 7.32 x 10^2 -> 2.23 x 10^0Improved
use DGCL instead of PANDA for brain network constructionbrain network construction · ABIDEreliability (ICC): 0.64 -> 0.71Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy