Data Sheet 1_The role of AI-driven decision support in immunotherapy for predicting pathol

248 patient records from May 2023 to December 2025, collected by Di Lu, focus on triple-negative breast cancer (TNBC) patients receiving neoadjuvant chemotherapy with or without pembrolizumab. The dataset was used to develop machine learning models for predicting pathological complete response (pCR). It is licensed under CC-BY-4.0 and was last updated in April 2026.

Use Cases

• Predicting pathological complete response (pCR) based on clinical variables like pembrolizumab use, age, and Ki67.
• Comparing treatment efficacy between pembrolizumab-chemotherapy and chemotherapy-alone groups.
• Developing interpretable machine learning models for clinical decision support using SHAP analysis.
• Training models on variables selected by LASSO regression, including SLD, N stage, and clinical stage.

Strengths

• Includes 248 patient records with a defined time range from May 2023 to December 2025.
• Data was used to build and validate eight machine learning models, with the MLP model achieving an ROC-AUC of 0.71.
• Propensity score matching was applied to balance baseline characteristics between treatment groups.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is known, but the specific variables and their distributions are not detailed in the metadata.
• Data may reflect temporal and selection bias inherent to a single retrospective study.

Provenance

Source

Di Lu via figshare

Collection Method

Retrospectively collected clinical data

Time Range

May 2023 to December 2025

Freshness

Last updated 2026-04 13 05:36:00.

Geography

null