Supplementary file 1_AI-driven spatial analysis of tumor-infiltrating lymphocytes predicts

32 hematoxylin-eosin-stained breast cancer sample images from a study by Hai Liang, last updated in 2026. The dataset underpins an AI pipeline for quantifying tertiary lymphoid structures and tumor-infiltrating lymphocytes to predict chemo-immunotherapy response. The predictive model based on spatial features achieved an AUC of 0.81.

Use Cases

• Train AI models for automated detection of tertiary lymphoid structures based on HE-stained images.
• Develop predictive models for therapy response based on lymphocyte spatial organization and proximity to tumor cells.
• Validate spatial analysis metrics like stromal lymphocyte density and mean minimum distances against clinical outcomes.
• Benchmark AI classifier performance against pathologist annotations for tumor-infiltrating lymphocyte quantification.

Strengths

• The TLS recognition model demonstrated substantial agreement with pathologists (κ=0.73).
• The TIL classifier achieved 0.92 accuracy.
• The spatial-feature-based prediction model achieved an AUC of 0.81 (95% CI: 0.76–0.86).
• Data is openly shared under a CC-BY-4.0 license.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The study notes external validation in larger cohorts is warranted.

Provenance

Source

Hai Liang via figshare.

Collection Method

Integrated AI pipeline for automated analysis of HE-stained breast cancer samples.

Time Range

null

Freshness

Last updated 2026-04-16 05:26:04; freshness should be verified.

Geography

null