This is the very first release of a new concept I have been wanting to try for a while: a new concept of article that evolves over time and with incremental updates of the associated registration algorithms.
This paper focuses on the block-matching algorithm for medical images registration and the core, behind the hood, developments I have made to it (with numerous students and post-docs) over the years. It cannot be self-contained but hopefully will contain stuff of interest to the readers.
The possibilities offered by a Git platform will be used to provide:
- releases when sufficiently important changes happen to the paper (new set of experiments, new methodological developments, etc.)
- a discussion forum opened for discussion on the paper and methodology. It can include either reviews of a part or all of the paper, discussion on new ideas related to the paper, etc.
Associated to this article, all developments proposed are available in the Anima software from the Empenn team. See at these two locations for trying it out:
Feel free to like (or dislike) this article just under this release text, comment, leave a constructive review, or just discuss on it in the discussion forum. Thanks for your input !