AI Paper Atlas

Research map / Detection and segmentation / CNN features

SpaCRD: Multimodal Deep Fusion of Histology and Spatial Transcriptomics for Cancer Region Detection

Shuailin Xue, Jun Wan, Lihua Zhang, Wenwen Min

AAAI 2026 · Detection and segmentation › CNN features · approach: Domain-specific segmentation and detection · 1 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
replace histology-pretrained UNI extractor with HIPT (ablation)SpaCRD · IDC cross-platformAUC: 0.891 -> 0.399Hurt
replace UNI with ResNet50 (natural-image pretrained) (ablation)SpaCRD · IDC cross-platformAUC: 0.891 -> 0.596Hurt
replace UNI with Swin-Tiny (ablation)SpaCRD · 10XHBC cross-platformAUC: 0.900 -> 0.689Hurt
use UNI instead of ResNet50 or Swin-Tiny (ablation)SpaCRD · STHBC cross-sampleAUC: 0.878 -> 0.929 (vs ResNet50)Improved
add bidirectional cross-attention (BCA) fusion (ablation)SpaCRD · HBC eleven datasetsAUC: 0.849 -> 0.923Improved
add category-regularized VAE (RVAE) (ablation)SpaCRD · HBC eleven datasetsAUC: 0.887 -> 0.923Improved
add whole VRBCA fusion network (ablation)SpaCRD · HBC eleven datasetsAUC: 0.815 -> 0.923Improved
add contrastive modality-alignment learning (ablation)SpaCRD · CRC (twelve datasets)AUC: 0.824 -> 0.869Improved
fuse histology with ST rather than histology only (ablation)SpaCRD · CRC (twelve datasets)AUC: 0.606 -> 0.869Improved
fuse ST with histology rather than ST only (ablation)SpaCRD · HBC eleven datasetsAUC: 0.832 -> 0.923Improved
add the fused variational loss (gamma 0.1 instead of 0.0) (ablation)SpaCRD · HBC eleven datasetsAUC: 0.772 -> 0.923 (gamma 0.0 to 0.1)Improved
set contrastive balance alpha to 0.5 instead of 0.0 or 1.0 (ablation)SpaCRD · HBC eleven datasetsAUC: 0.908 / 0.916 -> 0.923Improved

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