Clinical and MRI Radiomics for Ovarian Sex Cord-Stromal Tumor Classification

A multimodal nomogram integrating clinical features, MRI, and radiomics achieved an AUC of 0.945 for differentiating benign from malignant ovarian tumors. The dataset includes 123 pathologically confirmed tumors from 113 patients, used to develop and validate diagnostic models. Shifang Tan authored this research, last updated on 2026-05-25.

Use Cases

• Training machine learning models for tumor classification based on clinical parameters and MRI features.
• Developing combined diagnostic nomograms integrating radiomic signatures with traditional clinical predictors.
• Benchmarking the performance of radiomics models against conventional MRI analysis for ovarian cancer.
• Exploring feature importance for tumor malignancy using extracted radiomics from T2-weighted and diffusion-weighted imaging.

Strengths

• Includes 123 pathologically confirmed tumor samples, providing a ground truth for model training.
• Model performance is rigorously evaluated with AUC metrics and confidence intervals reported for training and validation cohorts.
• The combined nomogram demonstrated a high AUC of 0.945 in the training set, suggesting strong predictive potential.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is small at 84.1 KB, indicating limited scope.

Provenance

Source

figshare

Collection Method

Retrospective enrollment of patients with pathologically confirmed tumors.

Freshness

Last updated 2026-05-25 05:31:13; freshness should be verified.