Stroke Patient Data for Predicting 90-Day Functional Outcome After Low-Dose Alteplase

A retrospective study of 202 acute ischemic stroke patients receiving thrombolysis between August 2019 and August 2023, conducted by Huiru Chen. The dataset was used to develop machine learning models predicting 90-day functional outcomes based on clinical features like age, blood pressure, and NIHSS score.

Use Cases

• Developing a binary classifier for 90-day functional prognosis (mRS ≤2) based on clinical and imaging features.
• Comparing the performance of five machine learning algorithms (including Random Forest) for clinical risk stratification.
• Interpreting model predictions for key features like age, baseline NIHSS score, and hemorrhagic transformation using SHAP analysis.
• Validating a predictive model for poor outcomes in stroke patients treated with low-dose alteplase in an extended time window.

Strengths

• Dataset includes 202 patient records with a defined 7:3 training-validation split.
• Model development employed multiple evaluation metrics including AUC, DCA, and calibration curves.
• Key predictive features are explicitly listed, including age, NIHSS score, and imaging volumes.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data may reflect temporal and institutional bias inherent to the single retrospective study.

Provenance

Source

Huiru Chen via figshare.

Collection Method

Retrospective analysis of patient records.

Time Range

August 1, 2019 to August 31, 2023.

Freshness

Last updated 2026-05 08 05:55:42; freshness should be verified.

Geography

null