12,632,483
pareIT's method for turning mixed medical records into organised, verified, page-cited facts is protected by a U.S. patent, issued May 19, 2026.
The method, step by step
The patent protects a method for turning a stack of records from many providers into organised, linked medical data you can question. In the patent, "repair information" includes the treatments, procedures and medications performed on a patient. Here is the method in plain language, and where each step runs in pareIT.
Records from many providers, one file
Takes in records from many providers, created at different times, in structured and unstructured formats.
Reads page images and text together
Combines image-based and text-based analysis of every page, so scans, faxes and handwriting are read, not skipped.
Finds where each record begins
Looks at the page images to propose where separate documents were merged into one file.
Double-checks every boundary
Reads the text around each proposed boundary, across pages, to confirm the two sides really belong to different records.
Breaks each record into sections
Identifies the structure of each record from its layout and turns it into sections.
Labels and filters
Gives every record and section a category, and sets aside what is not needed for the analysis.
Extracts and links the medical data
Pulls out the medical data by type and links it across providers, onto one timeline.
Answers your questions
Lets you ask about the record and answers from the extracted data and its timeline.
"A technical improvement in the field of medical information extraction and presentation."
This is a plain-language summary. The patent's claims are the legal definition of what it covers.
License the technology.
Building in medical record review? Talk to us about licensing pareIT's patented technology.
Contact us about licensing →