Doctors in the Loop

Best medical annotation tools for Healthcare AI

About the series

Are you unsure which annotation tools are best suited for your healthcare project?

At Humans in the Loop, we offer annotation services for medical data such as X-rays, MRI and CT scans, microscopic imagery, ultrasound videos, and other 2D, 3D, and video data. Our global workforce of medical annotators comes from conflict-affected and displaced communities in various locations worldwide.

We are pleased to share our experts insights on the advantages and disadvantages of the top medical annotation tools.

The 3D Slicer segmentation tool is one of the most powerful and versatile tools available to medical professionals today. This open-source software is used to delineate regions and perform precise segmentation. Some of its benefits include:

  • DICOM standard interoperability for 2D, 3D and 4D images 
  • Image segmentation based on tissue density
  • AI model integrations, such as with NVIDIA’s MONAI

A key advantage of the tool is its integration with other medical imaging tools. This integration enables users to directly import and export medical images and segmentation files from other software applications, such as MRI and CT scan machines. However as an offline tool, it lacks quality control and management features that would allow you to track performance and easily manage all of your project’s tasks.

Our opinion :

We recommend it for 3D segmentation of CT scans as it has many modules which allow you to deal with different types of CT scans. The Slicer RT module is one of them that allows you to read and write DICOM Radiation Therapy objects (RT structure set, dose, image, plan, etc.) and provides tools for visualizing and analyzing them.

Encord is an annotation tool with specialized features for medical and healthcare data, which supports native DICOM and NIfTI image rendering with a PACS-style interface.

Its DICOM tool has been developed alongside clinicians and healthcare data scientists, delivering efficient functionalities and many useful features in its user interface which has configurable window presets. Additionally, it supports labeling videos of any length. This means any video can be uploaded, regardless of its format or length. Some other benefits include: 

  • Collaboration and custom QC workflow features
  • In-depth labeling protocols with nested classes
  • Supports multiplanar reconstruction, worklists and maximum intensity projection 

Our opinion :

Encord is an incredibly valuable tool for medical data annotation, given how versatile it is and how many different formats and annotations it supports. In addition, it’s suitable for large-scale projects with complex workflows and multiple annotators with consensus or QC labeling.

OHIF (Open Health Imaging Foundation) is a nonprofit organization that develops open-source software frameworks for medical image viewing, processing, and analysis. 

The OHIF annotation tool is a web-based application built on top of the OHIF viewer, which is an open-source tool for loading and viewing various medical imaging formats. The viewer can retrieve and load images from most sources and formats, render sets in 2D, 3D, and reconstructed representations, and enable the manipulation, annotation, and serialization of observations. Some of its benefits include:

  • 2D and 3D medical image viewing
  • Multiplanar Reconstruction 
  • Maximum Intensity Project 
  • Whole slide microscopy viewing
  • PDF and Dicom Structured Report rendering

Our opinion :

The OHIF annotation tool is a powerful and flexible tool that facilitates collaboration between annotators who can leave notes with additional clarifications for the labels and share annotations with colleagues. Highly recommended open source alternative to paid tools.

RedBrick is a SaaS annotation tool launched in 2021 with the goal to help healthcare AI teams annotate medical data more effectively. It offers annotation tools for CTs, MRIs, X-rays, etc., as well as comprehensive project management and quality control tools.

RedBrick AI also provides a web-based DICOM annotation tool with native support for DICOM medical images. It supports 2D and 3D data and allows for segmentation, classification, and vector annotations. The platform provides an intuitive and user-friendly interface, designed to be easy to use, even for those with limited technical experience. This ensures that medical professionals can spend less time navigating complex software. Some benefits include:

  • Segmentation with Brush, Pen, Contour tool, and Region growing
  • Smart interpolation and thresholding
  • Workforce metrics tracking for productivity and quality

Our opinion :

The Redrick medical annotation platform is one of the leading SaaS tools on the market and it provides a comprehensive solution for annotating medical data securely. Highly recommended for those looking for advanced project management and collaboration tools.

Supervisely released its medical interface in 2017 and since then has provided a well-known intuitive interface adapted to healthcare data.

It gives users the ability to view and manipulate volumetric medical images in multiple projections and slices with expert features for CT and MRI labeling in 2D or 3D. This is accompanied by Supervisely’s great features for labeler management, job queue management, and quality control. Some benefits include:

  • Navigation through slices in multiple projections
  • Perspective 3D view panel and multi-window layout
  • Multi-planar labeling features
  • Building 3D volumes from 2D figures

Our opinion :

We recommend it for beginners and simple projects as it provides an easily operated user interface and multi-planner labeling for DICOM annotation. It provides a variety of features that enable healthcare professionals to ensure that their medical image annotations are accurate. It also ensures ease of collaboration between labeling teams and managers.

V7 was launched in 2018 and shortly afterwards it released its full suite of medical annotation features for radiology, dentistry, pathology, etc.

V7 provides an auto-annotation feature for 3D medical images that speeds up your annotation time by using AI models to annotate your data. In addition, it supports custom labeling pipelines which incorporate human labeling with ML model input so as to speed up the annotation process. It’s FDA and HIPAA compliant. Some of its benefits include:

  • Consensus workflows between annotators or between a human and an AI model
  • Volumetric annotation for 3D data
  • Integrations with MLOps tools

Our opinion:

V7 is a powerful annotation tool, very intuitive and user-friendly. Its automation features are great for easy projects and high-quality datasets but we don’t recommend them for complex projects as they can’t always provide accurate annotations and adjusting the AI model responses may be more time-consuming than doing them from scratch. 

The right annotation team for your medical AI project

Get in touch with us and our team will assist you with finding the best solution

Leave a Comment

Your email address will not be published. Required fields are marked *