Carlo Cagnetta
Yordanka Velikova, Vanessa Gonzalez Duque, Felix Dülmer
Michael Panchenko
Chair of Computed Aided Medical Prociedures & Augmented Reality
School of Computation, Information and Technology
Technical University Munich
This repo contains the work of the Master Thesis of the author at the CAMPAR chair at TUM
To quickly set up the environment we advice using a conda env. The file ´´´requirements.txt´´´ contains a pip freeze of the current environment used for development.
This folder contains all code used for 2D image navigation in 3D MRI labels model.
- Notebooks: Used for function development. Contain logic explanation and use examples of functions. Fucntions developed in the Notebooks are then copied into ´´´.py´´´ files soto be re-used.
- 0_mri_visual: How to extract image information from 3D model. Tried scipy interpolation. OUTDATED
- 1_simple_clustering: Simple clustering with center-symmetric structure matrix using
scipy.ndimage.label. Also developed loss function and linear sweep proof of concept. - 2_DBSCAN_clustering: Better clustering algorithm using DBSCAN, tested on linear sweep loss.
- 3_slicing: Arbitrary linear sweep using simpleITK and euler matrix transformation.
- 4_environment: Showcase of how to get loss from an arbitrary slice.