Data Acquisition & De-Identification
Haystack integrates into a healthcare site's electronic health record (EHR), ingesting both structured and unstructured data. Major data sources include:
Provider notes
Lab results
Pathology
Scanned Results
Pharmacy Data
Unstructured data is ingested through our proprietary natural language processing (NLP) and optical character recognition (OCR) technology. During data ingestion, our programmatic de-ID engine redacts protected health information (PHI) with greater than 99% accuracy and without loss of clinical content. Our de-ID technology surpasses HIPAA Expert Determination standards, ensuring patient privacy is never compromised.
Clinical AI/NLP Patient Matching Engine
Meta-pragmatic clinical NLP on individual medical records
High accuracy clinical inferences generated from the longitudinal patient journey
Adaptable to any therapeutic area and study protocol
