Pearl

Pearl Second Opinion Image Analysis API

Partner/OEM-gated cloud AI surface behind Second Opinion. When an imaging partner uploads a bitewing, periapical, panoramic, or CBCT radiograph, Pearl's cloud computer-vision service analyzes it and returns detections - up to 18 findings per image (incipient/progressed caries, bone loss with millimeter and ratio measurements, calculus, periapical radiolucencies, defective margins, widened PDL, impactions) plus existing restorations (crowns, fillings, bridges, root canals, implants) - as labeled bounding boxes and metadata overlaid on the original image. There is no public, self-serve developer reference for this surface; endpoints are modeled from Pearl's product and integration descriptions.

GitHub

Documentation

📖
Documentation
https://hellopearl.com/products/second-opinion

Other Resources

🔗
Integrations
https://www.hellopearl.com/native-integration
🔗
APIsJSON
https://raw.githubusercontent.com/api-evangelist/pearl-dental/refs/heads/main/apis.yml

API entry from apis.yml

apis.yml Raw ↑ 
aid: pearl-dental:pearl-dental-second-opinion-image-analysis-api
name: Pearl Second Opinion Image Analysis API
tags:
- Image Analysis
- Radiographs
- Pathology Detection
- Partner Gated
- Modeled
image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
humanURL: https://hellopearl.com/products/second-opinion
properties:
- url: https://hellopearl.com/products/second-opinion
  type: Documentation
- url: https://www.hellopearl.com/native-integration
  type: Integrations
description: Partner/OEM-gated cloud AI surface behind Second Opinion. When an imaging partner uploads
  a bitewing, periapical, panoramic, or CBCT radiograph, Pearl's cloud computer-vision service analyzes
  it and returns detections - up to 18 findings per image (incipient/progressed caries, bone loss with
  millimeter and ratio measurements, calculus, periapical radiolucencies, defective margins, widened PDL,
  impactions) plus existing restorations (crowns, fillings, bridges, root canals, implants) - as labeled
  bounding boxes and metadata overlaid on the original image. There is no public, self-serve developer
  reference for this surface; endpoints are modeled from Pearl's product and integration descriptions.