Automatic Rib Segmentation in CT Data
Abstract
A supervised method is presented for the detection and segmentation of ribs in computed tomography ( ct ) data. In a first stage primitives are extracted that represent parts of the centerlines of elongated structures. Each primitive is characterized by a number of features computed from local image structure. For a number of training cases, the primitives are labeled by a human observer into two classes (rib vs. non-rib). This data is used to train a classifier. Now, primitives obtained from any image can be labeled automatically. In a final stage the primitives classified as ribs are used to initialize a seeded region growing process to obtain the complete rib cage. The method has been tested on 20 images. Of the primitives, 96.9% is classified correctly. The results of the final segmentation are satisfactory.
Cite
Text
Staal et al. "Automatic Rib Segmentation in CT Data." European Conference on Computer Vision, 2004. doi:10.1007/978-3-540-27816-0_17Markdown
[Staal et al. "Automatic Rib Segmentation in CT Data." European Conference on Computer Vision, 2004.](https://mlanthology.org/eccv/2004/staal2004eccv-automatic/) doi:10.1007/978-3-540-27816-0_17BibTeX
@inproceedings{staal2004eccv-automatic,
title = {{Automatic Rib Segmentation in CT Data}},
author = {Staal, Joes and van Ginneken, Bram and Viergever, Max A.},
booktitle = {European Conference on Computer Vision},
year = {2004},
pages = {193-204},
doi = {10.1007/978-3-540-27816-0_17},
url = {https://mlanthology.org/eccv/2004/staal2004eccv-automatic/}
}