67,830 record sets contain digitized physiological waveforms and time-series measurements for approximately 30,000 ICU patients. The data includes signals such as ECG, arterial blood pressure, respiration, and PPG, representing quasi-continuous recordings of a patient's vital signs throughout their ICU stay. A subset of this database is matched and time-aligned with the MIMIC-III Clinical Database records.
Use Cases
- Develop predictive models for patient deterioration based on continuous vital sign waveforms.
- Train signal processing algorithms for artifact detection and noise removal in ECG and ABP signals.
- Conduct multimodal research by aligning waveform data with clinical records from the matched subset.
- Analyze long-term physiological trends from recordings that can span several weeks.
Strengths
- Contains 67,830 record sets for approximately 30,000 ICU patients.
- Includes multiple physiological signal types such as ECG, ABP, respiration, and PPG.
- A subset is matched and time-aligned with detailed clinical records from MIMIC-III.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count for the primary time-series data is unknown, which may limit suitability assessment.
- Last update date is unknown; freshness unverified.
Provenance
- Source
- Benjamin Moody
- Collection Method
- Contains digitized signals and periodic measurements from ICU patient stays.