Home
Products
Services
About
Demo
The DeIdentifier Logo, a product from Carina.

Innovative De-identification & Data Management Platform

De-identification: A Complete Solution

DeIdentifier offers a robust and comprehensive way to remove Protected Health Information (PHI) across all DICOM RT modalities, while maintaining the research value of the data. It can perform header, burned-in text, and facial de-identification while ensuring full HIPAA compliance.
A graphic depicting DeIdentifier's date shifting feature for DICOM Headers.

Study Interval Retention

  • Automatically shift all dates within DICOM headers and clinical data using a patient-specific random number.
A graphic that illustrates DeIdentifier's header checking feature.

Redundancy Checks

  • Multi-stage secure checks during de-identification prevent unexpected PHI leaks from all DICOM headers.
2 different medical images that have burned-in text omitted from them.

Burned-in Text Removal

  • Text detection via natural language processing removes PHI text & retains key information, such as orientations.
Six different medical images of a face, with the first three being original medical data, and the final three being images that have been de-faced to remove any identifiable features.

Masked-Based Defacing

  • Our patented defacing algorithm generates a smooth mask to eliminate facial features with minimal image distortion, with negligible differences demonstrated for radiomics calculations and AI model training for internal structures*.
  • Our algorithm provides robust support for CT and MR, with extended compatibility for PET-CT, RT Image, RT Dose, and RT Structure Sets by applying the same mask.
  • Allows for fused multi-modality inputs with registration.

* Organs that are touching the mask, such as the eyes and nose, will be affected.

A report that has essential PHI removed, and shows highlights of how DeIdentifier interprets data.

Report De-identification

  • Text-based reports can be uploaded, added to a project, and automatically de-identified to remove PHI using DeIdentifier’s state-of-the-art large language model.

Multi-User Annotation Workflow

DeIdentifier’s 3D medical image visualizer, comprehensive suite of annotation features and multi-user annotation workflow provide the tools necessary for systematic and convenient data handling to support your clinical research.
An example study shown in DeIdentifier's viewer panel.

Web-Based Study Viewer

  • Access DeIdentifier 's study viewer to view 3D medical images from medical series, information regarding the study, along with a comprehensive set of features for creating and editing annotations.
  • DeIdentifier 's Comparison Viewer allows for longitudinal tracking across series.
  • Using DeIdentifier 's multi-user annotation workflow, reviewers can check and compare all user annotations across a single series, with built-in support for assessing inter-observer variability directly within the viewer.
An example study that has had its series parsed, and sorted into groups.

Automatic Study Parsing

  • Auto-parsing will group series from raw files uploaded to DeIdentifier based on the image type in the built-in viewer, eliminating the need to group them manually.
An image of DeIdentifier's user interface, where users can be assigned to certain studies.

Multi-User Annotation Assignment

  • DeIdentifier’s multi-user annotation workflow allows project administrators to assign users to specific tasks and studies for annotation.
  • Assign user-created profiles to studies within projects randomly or manually, and assigned accounts will only be able to view and annotate the studies that have been assigned to them.
  • Users can assign predefined labels to series and studies, in addition to leaving notes in order for administrators to review in the workflow process.
An image of DeIdentifier's user interface, showing annotations on a series from multiple users.

Built-in Annotation Tools

  • DeIdentifier offers full-body contouring support within the built-in study viewer.
  • Users can define regions of interest with manual annotations from scratch, or using the provided AI-driven Smart Interpolation tool.
  • DeIdentifier 's multi-user workflow allows project users to add annotations, labels and notes.
  • Upon completion, reviewers are able to view and adjust annotations from users.
An image of DeIdentifier's user interface, showing how custom models can be added and run on medical series.

Custom Model Support

  • Run supported user-built algorithms through DeIdentifier’s user interface.
  • Develop a model tailored for your research - from targeted ROI segmentation to calculating quantitative data within a specific series!
  • Choose from all compatible models and the series order preference when running, and the resulting output will be reflected directly in the built-in viewer.
Web-Based Study Viewer
Upload local data or batch retrieve available data from PACS remotely.
Automatic Study Parsing
Automatically sort and group a study's series for straightforward organization.
Multi-User Annotation Assignment
Configure de-identification projects to fit the scope of your research.
Built-in Annotation Tools
Use a variety of annotation tools to define and analyze areas of interest on de-identified slices.
Custom Model Support
Import user-developed scripts to run segmentation and analyzation tasks.

All-in-One De-identification & Data Management Solution

DeIdentifier offers a complete, user-focused solution with innovative features and an intuitive interface. Easily import data in multiple formats, customize project settings, and fine-tune and customize de-identification preferences - everything you need to take your work to the next level.
An image of DeIdentifier's user interface, illustrating remote study retrieval from a PACS server via DICOM Node connectivity.

Efficient Data Importing

  • DeIdentifier supports all formats of imaging (DICOM, NIFTI) and clinical data.
  • Batch retrieval uses CSV and REDCap files to allow users to pull data from PACS at once.
An image of DeIdentifier's user interface, showing an example project with studies.

Custom Project Configuration

  • Project-based data management allows for workflow configuration and customization for both de-identification and annotation-based projects.
  • Establish de-identification settings per project, including header and additional de-identification schema, export file format and fixed prefixes, and more!  
  • Assign users and establish predefined labels exclusive to each created project.
An image of DeIdentifier's user interface for uploading clinical non-imaging data to a project.

Automatic De-identification for Non-imaging Clinical Data

  • DeIdentifier’s project-based data management allows for automatic import and matching of clinical data, which will remain linked to the de-identified DICOM headers to preserve research value.
  • Smart Matching allows for one-click linking between clinical data and associated project studies when importing.
  • Linked clinical data will be de-identified congruently with all original study data.
A graphic illustrating DeIdentifier's export schema in relation to traditional Treatment Planning Systems.

Automatic Output Folder Organization

  • DeIdentifier allows for prefix naming configuration when creating new projects for exporting de-identified study folders.
  • Automatically organize study data chronologically with a patient/study/series hierarchy and permanent longitudinal tracking.
An image of DeIdentifier's user interface for sharing studies across external accounts.

Report De-identification

  • Easily share de-identified data between institutions while protecting essential PHI to spur collaborations on research projects.
  • Share projects through access tokens, or import project settings from external users.
  • Quickly toggle between local and shared studies on project page.
Efficient Data Importing
Upload local data or batch retrieve available data from PACS remotely.
Custom Project Configuration
Streamline data organization to manage de-identified data through user-defined projects.
Automatic De-identification for
Non-imaging Clinical Data
Efficient  management and simultaneous de-identification of clinical data.
Automatic Output Folder Organization
Organize data based on hierarchy and with customizable naming conventions.
Sharing & Collaboration
Project-based data management and sharing for both imaging & clinical data.

Thoroughly Secure All Your Data

DeIdentifier is installed locally within your institution’s firewall in order to prevent any external party from directly accessing the application and the PHI data.
A grahic of DeIdentifier's data protection and flow.
A table illustrating the differences between DeIdentifier and traditional Treatment Planning System features.

Contact Us

* indicates required
Copyright © 2025 - Carina Medical
[email protected]
+1 434-284-1073
DeIdentifier
QuantBrain - Ventricle
OncoAI Suite
INTContour
INTDose
Open-Source Tools
AI Development & 510(k) Validation
About Us
Events
News
Scientific Publications & White Papers
FCOI Policy
Privacy Policy
Copyright © 2024 - Carina Medical