Predictive Model for Symptomatic Lymphocele After Cervical Cancer Surgery, 221 Patients

221 patient records from a retrospective analysis at the Affiliated Hospital of North Sichuan Medical College, collected between January 2023 and September 2024. The dataset was used to develop and validate interpretable machine learning models for predicting symptomatic pelvic lymphocele, with an incidence rate of 19.9% in the cohort. Author Yiyue Wang published the supporting data on figshare under a CC-BY-4.0 license.

Use Cases

• Training interpretable ML models for surgical complication risk based on clinical characteristics and laboratory data.
• Identifying key predictive features for symptomatic lymphocele, such as diabetes and preoperative fibrinogen levels.
• Benchmarking model performance (e.g., KNN, XGBoost) using metrics like AUC and Brier Score on a specific clinical outcome.
• Applying SHAP analysis for feature importance explanation in a medical context.

Strengths

• Includes data from 221 patients with a defined outcome (44 with symptomatic lymphocele).
• Model performance metrics are provided, including a test set AUC of 0.832 for the best-performing KNN model.
• Key predictive features identified via SHAP analysis are listed in the description.

Limitations

• Row count and specific column definitions are unknown; field semantics must be inferred after download.
• Data is from a single hospital over a 20-month period, which may limit generalizability.
• The dataset is very small at 25.0 KB, indicating limited scope.

Provenance

Source

Affiliated Hospital of North Sichuan Medical College.

Collection Method

Retrospective analysis of clinical characteristics and laboratory data.

Time Range

January 2023 to September 2024.

Freshness

Last updated 2026-04-15 05:44:16; freshness should be verified.

Geography

Likely China (based on hospital location).