Barcelona SICU signals

Bedside monitors in intensive care units continuously generate high-resolution physiological signals that are rarely preserved systematically for research. This resource makes available to the community a set of multiparametric waveform recordings prospectively acquired in the surgical intensive care unit of Hospital Clínic de Barcelona (Spain). Data were captured with the open-source software VitalRecorder connected to Philips IntelliVue monitors, using a decentralized architecture without network transmission. Signals were recorded at their native sampling rates (electrocardiogram up to 500 Hz; plethysmography and invasive arterial pressure at 125 Hz; capnography, respiration and ventilator waveforms at 62.5 Hz; numeric trends at 1 Hz) and stored in the compressed binary .vital format. The set comprises 4,349 recordings (approximately 82 GB) totalling around 20,500 hours of monitoring over a twelve-month period. After privacy processing, 43 signal categories are published, all present in enough recordings to satisfy a k-anonymity threshold (k greater than or equal to 10, evaluated globally over the whole release, without temporal partitioning). The recordings, collected in pseudonymized form, are anonymized for this release: absolute timestamps, device and bed identifiers, patient data, and infrequent or potentially identifying signals are removed, leaving no re-identification key, and all times are relative to the start of each recording. Files are organized under a single folder with sequential names and are accompanied by per-recording metadata and documentation. The resource enables the development and validation of signal-based algorithms without the need for tabular clinical variables.

Verbatim fromPhysioNet· Licence: Creative Commons Attribution 4.0 International Public License