Fitting Skeletal Models via Graph-based Learning

Fitting Skeletal Models via Graph-based Learning

Abstract

Skeletonization is a popular shape analysis technique that models an object’s interior as opposed to just its boundary. Fitting template-based skeletal models is a time-consuming process requiring much manual parameter tuning. Recently, machine learning-based methods have shown promise for generating s-reps from object boundaries. In this work, we propose a new skeletonization method which leverages graph convolutional networks to produce skeletal representations (s-reps) from dense segmentation masks. The method is evaluated on both synthetic data and real hippocampus segmentations, achieving promising results and fast inference.

Citation

@INPROCEEDINGS{10635871,
  author={Gaggion, Nicolás and Ferrante, Enzo and Paniagua, Beatriz and Vicory, Jared},
  booktitle={2024 IEEE International Symposium on Biomedical Imaging (ISBI)}, 
  title={Fitting Skeletal Models via Graph-Based Learning}, 
  year={2024},
  volume={},
  number={},
  pages={1-4},
  keywords={Measurement;Image segmentation;Shape;Fitting;Neural networks;Training data;Skeleton;Geometric learning;Skeletal representations;Shape analysis;Graph-based neural networks},
  doi={10.1109/ISBI56570.2024.10635871}}