Automated Registration of Cone-Bean Computed Tomography scans - maintenance

[Jeanneclre]

Draft Status

Ready - team will start page creating immediately

Category

Other

Presenter Location

Online

Key Investigators

• Jeanne Claret (University of Michigan,USA)
• Gaëlle Leroux (UoM, USA)
• Eduardo Duarte Caleme (UoM, USA)
• Claudia Mattos (UoM, USA)
• Lucia Cevidanes (UoM, USA)
• Juan Prieto (University of North Carolina,USA)

Project Description

The Automated Registration tool, AREG, was first presented at the NA-MIC project week #39. It aims to reduce the sources of error in the 3D image processing workflow by automating the orientation and registration of 3D Cone-Beam Computed Tomography. These methods combine classical algorithmic approaches and AI-based models trained and tested on de-identified CBCT volumetric images.

The registration method is based on an automatic tool, AMASS, available in the extension SlicerAutomatedDentalTool, to perform a segmentation of the different regions of reference used for the regional voxel-based registration

The different methods for automatic orientation and registration of 3D CBCT scans rely on a combination of algorithmic and deep-learning techniques to perform both the orientation and the registration automatically. It also uses work that our group of researchers has already developed. Our Python-based algorithm and requires multiple libraries for the different image-processing tasks accomplished throughout the proposed method: SimpleITK, VTK, SimpleElastix. To implement these tools, we also used the Medical Open Network for Artificial Intelligence (MONAI) library, which is a PyTorch-based framework for medical image analysis. MONAI offers several advantages for our work, such as high performance, modularity, and interoperability with other libraries.

Objective

1. Maintain the code to make it work properly on the new version of Slicer

Approach and Plan

1. Find the issue by testing
2. Correct the problem

Progress and Next Steps

1. We found that AREG is not working with the ITK version 5.4.rc2 and we suppose it is the reason why it is not working with itk-elastix 0.19.1.
2. Need to add library versions to the code according to best practices for Slicer Extensions.

Illustrations

Comparison between the current and the proposed workflow

Different regions of reference (comparison between the full segmentation and the mask)

Example of Cranial Base Registration

Background and References

• AMASS and AREG

[github-actions[bot]]

Project Page Pull Request Creation

:white_check_mark: COMPLETED: See https://github.com/NA-MIC/ProjectWeek/pull/866