MITK v2023.12

Updated 142 Days AgoPublic

Actions

Previous changelog • Next changelog

News for MITK Workbench users

For MITK v2023.12, we focused mainly on segmentation and registration. In the past 6-7 months we resolved about 130 tasks in over 440 commits. We are already excited about your feedback on many improvements and features.

Segmentation

TotalSegmentator v2

The TotalSegmentator segmentation tool was updated to version 2 of TotalSegmentator. It now supports up to 117 classes for CT images. See the following official image from https://github.com/wasserth/TotalSegmentator for an overview of all new classes:

This is an example video of a low-resolution CT image segmented with the TotalSegmentator tool in the MITK Workbench v2023.12, followed by a conversion to smoothed meshes, and finally animated and recorded with the Movie Maker plugin:

Segment Anything Model (SAM)

We also introduce a new AI-powered tool for bio-medical images based on the popular Segment Anything Model from Meta. The interactive tool produces high-quality masks from a few seed points.

Registration

New registration algorithms

Two new 3d-3d registration algorithms were added to MITK v2023.12:

slabbedHead

A multimodal image registration algorithm optimized to robustly handle the registration of "slabbed" head images (i.e. MRI sequences where field of view is sacrificed to lower exposition time, while still providing high in-plane resolution) onto complete head images. The algorithm is more reluctant to make out-of-plane rotations, as there is less driving information available and the probability for errors is higher.

HeadNeck

A multimodal image registration algorithm optimized to robustly handle the registration of a head image with an image that also contains the neck or upper body parts. In its default settings the algorithm automatically decides which image is which based on the extent of the z dimension.

New registration command-line apps

Two new registration command-line apps were added to MITK v2023.12:

MitkMatchImage

Use registration algorithms provided by MITK from the command-line. It is the command-line equivalent to the MatchPoint Registration plugin.

MitkMapImage

Use registration results to map an image into the target space of the registration. It is the command-line equivalent to the MatchPoint Mapping plugin.

DICOM

The official release installers of MITK v2023.12 now include the following diagnostic tools:

MitkDICOMVolumeDiagnostics command-line app

Offers the DICOM volume analysis logic used by MITK as a command-line app. The app prints a JSON-based report that, among others, contains the information which reader MITK would use, which volumes were detected, and how many slices and frames these volumes have as well as which files contribute to each volume respectively.

DICOM Inspector plugin

Helps to inspect all DICOM contact information fetched by MITK. The handling is similar to the already known Properties plugin. If a selected data node contains DICOM information, it will be shown in the table of the new plugin. This information is temporally and spatially resolved, e.g. if you scroll through the slices or through time (in dynamic data sets) you will see the instance UID changing, always indicating the value that belongs to the crosshair position.

News for developers

See the past two changelogs starting from 2023 Week 46 (Mid November) for a comprehensive, developer-centric overview of changes.

Renaming of command-line apps

Some apps have been renamed, for example:

• MitkMapImageMiniApp MitkMapImage
• MitkMatchImageMiniApp MitkMatchImage
• MitkStitchImageMiniApp MitkStitchImage

API change of mitk::BaseProperty

See 2023 Week 48 (Very Late November) for a migration guide for the changed API of mitk::BaseProperty, which allows to (de)serialize properties and property lists in JSON format.

🛠 Third-party dependency changes

The following table shows a complete list of changed third-party dependencies. Notable changes are:

• ITK v5.3.0: You may experience a few compile errors in your code since a ITK had a few subtle API changes (see our fixes), in particular regarding...
  • points and vectors, resp. their conversion to each other or from plain numbers.
  • class macros in case you define new classes in the itk namespace.

🔥 Disclaimer for API-breaking changes

We discontinued the extensive listing of API-breaking changes as the vast majority of them are straight forward to resolve or do not affect the absolute majority of developers at all. The ratio between the time and effort spent for writing these reports and actual developer feedback turned out to be greatly imbalanced.

Complex or critical API changes will still be accompanied by migration guides. For example, see 2023 Week 48 (Very Late November) for a migration guide for the changed API of mitk::BaseProperty.

In case you experience any trouble while migrating to the latest version of MITK, please do not hesitate to contact us.

Previous changelog • Next changelog