AI Data Mapping (BETA)

AI Data Mapping allows you to output AI analysis into multiple fields in a collection section. 

Like AI Fields, the AI extracts and organizes information from documents attached to the input file attachment field. However, data mapping allows you to map the output into multiple fields in a collection section. With AI Data Mapping, AI output can be used to output different pieces of the analysis to different fields, automatically creating new collection items with their information filled. 

Change to AI Mapping Setup

If you’ve used AIFields, you may be familiar with adding the AI analysis to the output field (a text field), and then linking back to the input field (a file attachment field).

Both AI Fields and AI Data Mapping are now mapped from the input field (the file attachment field). This change allows you to map the input field to a multitude of different fields at once, rather than manually creating and linking each output field to the input one at a time. 

AI Fields and AI Data Mapping can both be toggled on under the AI Field Setup section of input field’s modal.

Step-by-Step Setup

AI Data Mapping is set up in the Customs Editor.

Create Your Destination Section and Input Field

1. If needed, create your destination section. Ensure the collection section where you want to output the data is created. You do not need to add fields, as they will be created during data mapping. (However, if you already have existing fields, you will be able to map to them.)
2. Create the input field. Add a file attachment field to be the input field. This input field can be in any collection section.
   Note: Data mapping is available on both single file attachment fields and multiple file attachment fields. If you are planning to run the AI analysis on multiple files, make sure that your input field is a multiple file attachment field.

Turn on Data Mapping

1. Add data mapping to the input field. Click Enable Data Mapping to turn it on and see the setup options.
2. Choose the AI prompt. Choose the AI analysis type you’d like to use to generate the output. Currently, the Medical Chronology analysis is the only AI analysis type available for data mapping.
3. Choose the destination section. In the Destination Section box, choose the collection section you choose for the output.

Field Mapping Assignments

The Medical Chronology analysis type outputs a list of pieces of information. View the full list in the subsequent section. You can toggle a field mapping assignment on to include that information from the output.

By default, data mapping uses auto-linking to create a new field in the destination section for each piece of information in the list. The new fields will be typed according to their data—“date of service” to a data field, “provider” to a single line text field, and so on. 

To map a piece of information to an existing field, hover on a selected field and click the edit icon. Select the field where you’d like to map the data (the field must be in the destination section). 

To return to auto-linking the field, click Auto-link.

Create

When you’re finished field mapping, click Create. Once you create it, all the selected new fields will be created in the destination section. You can view and edit them in that section in the Customs Editor as usual. Fields being used as AI input or output (for both AI Fields and AI Data Mapping) will be marked accordingly in the Customs Editor. 

Medical Chronology Analysis

Currently, AI Data Mapping allows for the analysis type Medical Records - Medical Chronology. This analysis type is meant to review a medical record. For each treatment date, the AI returns the following information. In AI Data Mapping, each piece of information can be mapped to a different field.

• Date of service
• Provider
• Facility
• Reason for Visit
• History of the present illness
• Subjective
• Chief Complaint
• Pain score
• Objective
• Vital signs
• Testing
• Imaging
• Assessment 
• Diagnoses
• ICD-10-CM Codes
• Treatment
• Medications prescribed
• CPT Codes
• Plan
• Follow ups or referrals ordered
• Prognosis

AI Data Mapping may expand to similar analysis types—that is, analysis types that find a list of pieces of information that would be useful to map to different fields.

Output

Once AI Data Mapping is set up, whenever a file is added to the input field, the AI will run analysis. When it’s done, the mapped fields will be filled. 

Every time a new document is uploaded to the multiple file attachment field, the AI analysis will run, and a new collection item will be created with the mapped fields filled.