Wound Vision

Wound Vision is an iOS clinician-assist application. It uses the iPhone camera, LiDAR, or TrueDepth data to record a wound. It estimates wound depth and volume. It also provides rule-based graft-size information for documentation.

Warning

Wound Vision is not a medical device. It is not validated for diagnosis or treatment. Do not use its output as the only basis for a clinical decision. Confirm each measurement with an approved clinical method.

Screenshots

Screenshots will use synthetic demonstration data. They will not contain protected health information.

The screenshot uses an empty simulator data store. It contains no patient data.

Main functions

• Capture an image with the rear camera or the TrueDepth camera.
• Use LiDAR data on a supported iPhone.
• Mark or correct the wound perimeter.
• Estimate depth, area, and volume on the device.
• Save a local capture history.
• Export a report or a CSV file.
• Process data without a remote inference service.

Privacy

Wound Vision processes images on the device. The application does not require an API key. The application does not upload a capture to a server.

The application stores captures in its Application Support directory. It uses complete file protection for stored capture files. The current prototype uses pseudonymous patient identifiers.

Review the privacy and data-retention requirements of your organization before you use the application with real patient data.

Requirements

• macOS with Xcode 16 or later
• XcodeGen
• iOS 18 or later
• An Apple development team for installation on a physical device
• An iPhone with TrueDepth for the front-camera workflow
• An iPhone Pro with LiDAR for the LiDAR workflow

The simulator can build the application. The simulator cannot reproduce all camera, TrueDepth, or LiDAR functions.

Included model

The repository includes WoundSegmentation.mlpackage. The application uses this Core ML model for on-device wound segmentation.

The root directory also can contain training checkpoints and exported model archives. These files are development output. Git does not include them. You do not need them to build or run the application.

Install the development tools

1. Install Xcode from Apple.
2. Install XcodeGen.

brew install xcodegen

Generate the Xcode project

1. Clone the repository.
2. Open a terminal in the repository.
3. Generate the project.

xcodegen generate

This command creates WoundVision.xcodeproj from project.yml.

Configure code signing

1. Open WoundVision.xcodeproj in Xcode.
2. Select the WoundVision target.
3. Open Signing & Capabilities.
4. Select your development team.
5. Change the bundle identifier if Xcode reports a conflict.

Do not commit a signing certificate, provisioning profile, or private key.

Build the application

Run this command for a simulator build:

xcodebuild \
  -project WoundVision.xcodeproj \
  -scheme WoundVision \
  -destination 'generic/platform=iOS Simulator' \
  -configuration Debug \
  build

Run the tests

Select an installed simulator name. Then run this command:

xcodebuild \
  -project WoundVision.xcodeproj \
  -scheme WoundVision \
  -destination 'platform=iOS Simulator,name=iPhone 16 Pro' \
  test

Project structure

WoundVision/
  App/          Application entry point and dependencies
  Data/         Local storage and repositories
  Domain/       Measurement and recommendation logic
  Features/     User workflows and screens
  Resources/    Assets, thresholds, and the Core ML model
  Support/      Device checks, logging, and utilities
  UI/           Shared visual components
WoundVisionTests/
project.yml

Security

Read SECURITY.md before you report a vulnerability. Do not include patient data, a private capture, an API key, or a signing file in an issue.

License

Wound Vision uses the PolyForm Noncommercial License 1.0.0. It is source-available software. It is not open source under the OSI definition. Business use requires written permission from One Man Labs.

Third-party components and model technology can have separate terms. Read THIRD_PARTY_NOTICES.md.