Structured data enables more informed decision making

Schedule Demo

Leveraging the cloud, computer vision, machine learning and natural language processing to efficiently abstract and structure data from faxes and clinical notes.

Data ingestion, especially from unstructured sources, is a headwind to realize the advantages of precision health. Unfortunately, healthcare is fraught with poor quality printed documents and faxes that need to be structured to be actionable. Typically, they are either scanned into an EMR as a file attachment, transcribed by expensive abstractors or even worse, end up as paper copies in filing cabinets.

We developed PrecisionOCR to easily and accurately unlock the value of this data. The PrecisionOCR solution uses custom Optical Character Recognition technology (OCR) combined with artificial intelligence algorithms to turn faxes and scanned documents into searchable, structured data.

PrecisionOCR is available as a standalone application where you can simply upload documents for a low per-page price or fully integrated into the Precision Health Cloud so data from these documents can be combined and analyzed with your other data sources.

PrecisionOCR is easy to use, cost effective and highly accurate. Even the messiest and illegible documents can be interpreted. Procedures, observations, medications and interventions are automatically highlighted. The contents of the document can easily be searched. Once the data is structured, it is converted to FHIR resources that can be used in the Precision Health Cloud as well as integrated back into your EMR systems.

PrecisionOCR Features:

  • Self-signup including a BAA (Business Associate Agreement) and no upfront costs.
  • All data is stored in LifeOmic’s HIPAA compliant, HITRUST CSF and SOC 2 Certified cloud platform. You can delete your data at any time.
  • Upload your documents via your web browser.
  • Leverage proprietary computer vision denoising algorithm to reduce visual artifacts and improve OCR.
  • Machine learning based document classifier trained using MIMIC-III data set. Recognizing that end users will have unique document classification needs, we have provided functionality to enable, maintain and grow custom document classifications.
  • Machine learning based OCR service to identify, understand, and extract data from forms and tables.
  • Machine learning based spellcheck trained using MIMIC-III data set.
  • Automatically surface a list of medications, conditions, procedures and observations.
  • Leverage public ontologies including LOINC and SNOMED for observations and procedures.
  • Define and maintain domain specific keywords to be surfaced within the document or clinical note.
  • Ability to search and filter across uploaded documents by keyword.
  • Ability to organize documents into folders.
Schedule Demo

Using PrecisionOCR in the Precision Health Cloud you can:

  • Automatically index and organize by patient (by name or MRN)

  • Automatically suggest FHIR resources based on keywords for indexing data from the document, making search functionality more robust and flexible

  • Leverage custom ontologies for searching & filtering

  • Use PrecisionOCR directly in the Precision Health Cloud web app.

Schedule Demo

Note: All data is synthetic and does not contain protected health data