Data and code: Mining routine ultrasound reports to re-examine biopsy burden in BI-RADS category 4a (consequence-weighted dual endpoints)

Aggregate figure- and table-level source data (40 CSV files, summary statistics only) and the report-text extraction and quality-control code (features_v2.py and QC scripts) for the manuscript on consequence-weighted dual endpoints in BI-RADS category 4a in a biopsy-confirmed breast ultrasound population. No patient-level records, ultrasound application numbers, dates, or free-text pathology are included. Data are released under CC BY 4.0; the code files under code/ are additionally licensed under the MIT License (see LICENSE-code.txt).