Saeed Mohagheghi1, Amir Hossein Foruzan2. 1. Department of Biomedical Engineering, Engineering Faculty, Shahed University, Tehran, Iran. 2. Department of Biomedical Engineering, Engineering Faculty, Shahed University, Tehran, Iran. a.foruzan@shahed.ac.ir.
Abstract
PURPOSE: Convolutional neural networks (CNNs) have obtained enormous success in liver segmentation. However, there are several challenges, including low-contrast images, and large variations in the shape, and appearance of the liver. Incorporating prior knowledge in deep CNN models improves their performance and generalization. METHODS: A convolutional denoising auto-encoder is utilized to learn global information about 3D liver shapes in a low-dimensional latent space. Then, the deep data-driven knowledge is used to define a loss function and combine it with the Dice loss in the main segmentation model. The resultant hybrid model would be forced to learn the global shape information as prior knowledge, while it tries to produce accurate results and increase the Dice score. RESULTS: The proposed training strategy improved the performance of the 3D U-Net model and reached the Dice score of 97.62% on the Sliver07-I liver dataset, which is competitive to the state-of-the-art automatic segmentation methods. The proposed algorithm enhanced the generalization and robustness of the hybrid model and outperformed the 3D U-Net model in the prediction of unseen images. CONCLUSIONS: The results indicate that the incorporation of prior shape knowledge enhances liver segmentation tasks in deep CNN models. The proposed method improves the generalization and robustness of the hybrid model due to the abstract features provided by the data-driven loss model.
PURPOSE: Convolutional neural networks (CNNs) have obtained enormous success in liver segmentation. However, there are several challenges, including low-contrast images, and large variations in the shape, and appearance of the liver. Incorporating prior knowledge in deep CNN models improves their performance and generalization. METHODS: A convolutional denoising auto-encoder is utilized to learn global information about 3D liver shapes in a low-dimensional latent space. Then, the deep data-driven knowledge is used to define a loss function and combine it with the Dice loss in the main segmentation model. The resultant hybrid model would be forced to learn the global shape information as prior knowledge, while it tries to produce accurate results and increase the Dice score. RESULTS: The proposed training strategy improved the performance of the 3D U-Net model and reached the Dice score of 97.62% on the Sliver07-I liver dataset, which is competitive to the state-of-the-art automatic segmentation methods. The proposed algorithm enhanced the generalization and robustness of the hybrid model and outperformed the 3D U-Net model in the prediction of unseen images. CONCLUSIONS: The results indicate that the incorporation of prior shape knowledge enhances liver segmentation tasks in deep CNN models. The proposed method improves the generalization and robustness of the hybrid model due to the abstract features provided by the data-driven loss model.
Keywords:
3D liver segmentation; Convolutional neural network; Deep learning; Prior knowledge
Authors: Tobias Heimann; Bram van Ginneken; Martin A Styner; Yulia Arzhaeva; Volker Aurich; Christian Bauer; Andreas Beck; Christoph Becker; Reinhard Beichel; György Bekes; Fernando Bello; Gerd Binnig; Horst Bischof; Alexander Bornik; Peter M M Cashman; Ying Chi; Andrés Cordova; Benoit M Dawant; Márta Fidrich; Jacob D Furst; Daisuke Furukawa; Lars Grenacher; Joachim Hornegger; Dagmar Kainmüller; Richard I Kitney; Hidefumi Kobatake; Hans Lamecker; Thomas Lange; Jeongjin Lee; Brian Lennon; Rui Li; Senhu Li; Hans-Peter Meinzer; Gábor Nemeth; Daniela S Raicu; Anne-Mareike Rau; Eva M van Rikxoort; Mikaël Rousson; László Rusko; Kinda A Saddi; Günter Schmidt; Dieter Seghers; Akinobu Shimizu; Pieter Slagmolen; Erich Sorantin; Grzegorz Soza; Ruchaneewan Susomboon; Jonathan M Waite; Andreas Wimmer; Ivo Wolf Journal: IEEE Trans Med Imaging Date: 2009-02-10 Impact factor: 10.048
Authors: Konstantinos Kamnitsas; Christian Ledig; Virginia F J Newcombe; Joanna P Simpson; Andrew D Kane; David K Menon; Daniel Rueckert; Ben Glocker Journal: Med Image Anal Date: 2016-10-29 Impact factor: 8.545