This repo contains a sample script for the ensemble methods explained in article "Basic Ensembles of Deep Learning Models for Liver Segmentation from CT Images". For detailed explanations, please refer to the article (currently in review).
The code was written in MATLAB. ensemleDeepModels_MAIN.m is the main script that users need to execute. There are evaluation of four individual segmentation methods and implentation of five different ensemble methods as well as their evaluations in the script. Data is provided from CHAOS challenge dataset, CT Set 2.
Besides of all files in this repo, it is necessary to download deepModelResults.mat (143 MB) from the provided link. This file stores probability maps coming from four individual Deep Models for CHAOS CT Set 2. These models are:
-
DeepMedic: K. Kamnitsas, E. Ferrante, S. Parisot, C. Ledig, A. V. Nori, A. Criminisi et al., “DeepMedic for brain tumor segmentation,” in Lecture Notes in Computer Science, vol. 10154 LNCS. Springer, Cham, oct 2016, pp. 138–149. https://link.springer.com/chapter/10.1007/978-3-319-55524-9_14
-
Dense V-Networks: E. Gibson, F. Giganti, Y. Hu, E. Bonmati, S. Bandula, K. Gurusamy et al., “Automatic Multi-Organ Segmentation on Abdominal CT with Dense V-Networks,” IEEE Transactions on Medical Imaging, vol. 37, no. 8, pp. 1822–1834, aug 2018. https://ieeexplore.ieee.org/document/8291609/
-
U-net: O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in Lecture Notes in Computer Science, vol. 9351. Springer, Cham, 2015, pp. 234–241. https://link.springer.com/chapter/10.1007%2F978-3-319-24574-4_28
-
V-net: F. Milletari, N. Navab, and S. A. Ahmadi, “V-Net: Fully convolutional neural networks for volumetric medical image segmentation,” in Proceedings - 2016 4th International Conference on 3D Vision, 3DV 2016. IEEE, oct 2016, pp. 565–571. http://ieeexplore.ieee.org/document/7785132/
The ensemble combination methods are: majority vote, average, product and min/max (Note that min and max combiners are identical for two-class problems).
- Ludmila I. Kuncheva, Combining Pattern Classifiers: Methods and Algorithms: Second Edition. Wiley-Interscience, 2014, ISBN:0471210781, vol. 9781118315. https://onlinelibrary.wiley.com/doi/book/10.1002/9781118914564
The evaluation is handled by CHAOS challenge metrics. For more information about ;
- CHAOS Challenge: https://chaos.grand-challenge.org/
- CHAOS Evaluation Code: https://github.com/emrekavur/CHAOS-evaluation
- CHAOS Dataset: https://zenodo.org/record/3367758