When using the Intelligent MedRecords (IMR) solution, note the following considerations:
File Types and Formats
- Because OCR quality can affect classification and extraction, it is recommended that documents have a minimum resolution of 300 DPI.
- IMR can only accept PDF and image file types (including TIFF, JPG, JPEG, and PNG). Content Federation Services (CFS) can convert other file types (such as .txt, .doc, and .docx) to PDF when uploading the documents to the Content Innovation Cloud. The preferred and expected way for IMR to receive documents is through capture in OnBase and sending through CFS. However, for testing and troubleshooting, IMR allows you to upload documents manually by clicking the Start Process button and submitting the Upload Batch form. Please note that although the Upload Batch form allows documents of any file type to be uploaded, you should select only documents with PDF or image file types. Other file types may cause processing errors or exceptions.
- IMR returns all documents as PDF.
System Limits
Version 1.0 of IMR can support the following system limits:
- Maximum of 100 classes (Document Types) configured
- Maximum of 140 pages single file size for classification and extraction
- Maximum of 500 documents per batch for document separation
- Maximum of 1000 characters in Document Class Description (AI prompt)
- Maximum of 1000 characters in Field Configuration (AI prompt)
Customer Processing
Customers will need to complete indexing if values are not populated.
OnBase
- Customers have or will have an OnBase environment in place (Foundation 24.1 or later) before implementing IMR.
- IMR does not support OnBase Client (Thick Client) scanning. Customers must be able to send documents to a Capture Process Designer (CPD) enabled scan queue.
- IMR does not support direct HL7 sending. Customers must have a Workflow-based HL7 sending process in place.