Sample preparation
This protocol is extracted from research article:
Respiratory deformation registration in 4D-CT/cone beam CT using deep learning
Quant Imaging Med Surg, Feb 1, 2021; DOI: 10.21037/qims-19-1058

For a pair of moving image M and target image T with their deformation vector field ϕ, a patch pair {PM(u), PT(u)} and the displacement vector ϕ(u) = (dx,dy,dz) at the same position u were extracted, where PM(u) is the patch of moving image, PT(u) is the patch of the target image, and ϕ(u) is the displacement vector registered by VelocityAI at position u. A sample is represented as {PM(u), PT(u)lϕ(u)}. The characteristics of the sample were determined by the patch center and the patch size. The centers of the patch pair determine the location of the patch, and the patch size determines the range of features included in the patch pair.

The patch center points were chosen on a 3-dimensional uniform grid across the lung. The density of the grid determines the number of samples. In order to ensure sufficient training samples, the density was chosen such that the number of patch pair samples is around 20,000. The center points sampling density was designed to be denser along the superior-inferior direction than the anterior-posterior and medial-lateral directions because the magnitude of the deformation field along superior-inferior direction is generally larger than the other two directions. The patch size was chosen to be large enough to cover the motion range of the anatomical structures within the patch.

Note: The content above has been extracted from a research article, so it may not display correctly.



Q&A
Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.



We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.