CT-Based Radiogenomic Prediction of NK Cytotoxicity Biomarkers in Kidney Cancer

143 patients with clear cell renal cell carcinoma (ccRCC) contributed preoperative CT images, with transcriptomic data from 538 TCGA-KIRC tumor samples. The dataset, authored by Xinwei Ma and last updated in June 2026, supports a framework linking 3,176 radiomic features per tumor to the expression of immune biomarkers ICAM1 and RAET1E. External validation was performed using tissue microarrays from 26 independent ccRCC cases.

Use Cases

• Train machine learning models to predict ICAM1 expression levels based on CT radiomic features.
• Validate radiogenomic associations for RAET1E as a biomarker related to natural killer cell cytotoxicity.
• Perform survival analysis linking imaging-derived biomarker estimates to patient outcomes in kidney cancer.

Strengths

• Includes data from 143 patients with histologically confirmed ccRCC and transcriptomic data from 538 TCGA-KIRC samples.
• Extracted 3,176 radiomic features per tumor spanning seven feature classes and eight image filters.
• External validation performed on an independent cohort of 26 ccRCC cases using immunohistochemistry.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small at 16.7 KB, suggesting it contains summary or model results rather than raw feature data.

Provenance

Source

figshare

Collection Method

Radiomic features were extracted from manually segmented 3D tumor volumes on preoperative contrast-enhanced CT images.

Freshness

Last updated 2026-06-03 05:48:17; freshness should be verified.