Lung Adenocarcinoma H&E Slides with TIL Spatial Patterns for Immune Subtyping

A 2026 study by Xia Li developed an interpretable prediction model for lung adenocarcinoma immune subtyping. The work used transcriptomic data from 503 TCGA LUAD patients and automated annotation of H&E whole-slide images to quantify spatial distribution patterns of tumor-infiltrating lymphocytes. The model was validated in internal and external cohorts, achieving AUCs of 0.839 and 0.927, respectively.

Use Cases

• Training image-based models for immune subtyping based on TIL spatial distribution patterns.
• Developing interpretable prediction tools for tumor immune microenvironment evaluation.
• Linking morphological phenotypes from H&E slides to molecular immune subtypes.
• Validating computational pathology models using internal and external cohorts.

Strengths

• Model validation reported strong performance metrics, including a 95.09% Dice score for tissue segmentation and an AUC of 0.927 in external validation.
• The study integrates multiple data modalities: transcriptomic analysis, image-based TIL quantification, and immunohistochemical confirmation.
• The dataset and methodology are fully described in a CC-BY-4.0 licensed document, supporting open science.

Limitations

• The primary data file is a 341.0 KB DOC document; the underlying image and tabular datasets are not directly accessible.
• Row count and column-level documentation for any associated data are unknown, limiting suitability assessment.
• Freshness should be verified; the last update was 2026-05-08 05:40:47.

Provenance

Source

figshare, authored by Xia Li.

Collection Method

Research study integrating transcriptomic analysis from TCGA and automated annotation of H&E whole-slide images.

Time Range

Study published in 2026; specific temporal coverage of patient data is not stated.

Freshness

Last updated 2026-05 08.

Geography

null