Frangi-Net: A Neural Network Approach to Vessel Segmentation

Fu W, Breininger K, Schaffert R, Ravikumar N, Würfl T, Fujimoto JG, Moult EM, Maier A (2018)

Publication Language: English

Publication Type: Conference contribution, Original article

Publication year: 2018

Publisher: Springer Vieweg, Berlin, Heidelberg

City/Town: Berlin, Heidelberg

Pages Range: 341-346

Conference Proceedings Title: BildVerarbeitung für die Medizin (BVM) 2018

Event location: Erlangen

ISBN: 978-3-662-56537-7

URI: https://arxiv.org/abs/1711.03345

DOI: 10.1007/978-3-662-56537-7_87

Abstract

In this paper, we reformulate the conventional 2-D Frangi vesselness measure into a pre-weighted neural network (“Frangi-Net”), and illustrate that the Frangi-Net is equivalent to the original Frangi filter. Furthermore, we show that, as a neural network, Frangi-Net is trainable. We evaluate the proposed method on a set of 45 high resolution fundus images. After fine-tuning, we observe both qualitative and quantitative improvements in the segmentation quality compared to the original Frangi measure, with an increase up to 17% in F1 score.

Authors with CRIS profile

Involved external institutions

Massachusetts Institute of Technology (MIT) United States (USA) (US)

How to cite

APA:

Fu, W., Breininger, K., Schaffert, R., Ravikumar, N., Würfl, T., Fujimoto, J.G.,... Maier, A. (2018). Frangi-Net: A Neural Network Approach to Vessel Segmentation. In Andreas Maier, Thomas M. Deserno, Heinz Handels, Klaus H. Maier-Hein, Christoph Palm, Thomas Tolxdorff (Eds.), BildVerarbeitung für die Medizin (BVM) 2018 (pp. 341-346). Erlangen, DE: Berlin, Heidelberg: Springer Vieweg, Berlin, Heidelberg.

MLA:

Fu, Weilin, et al. "Frangi-Net: A Neural Network Approach to Vessel Segmentation." Proceedings of the BildVerarbeitung für die Medizin (BVM) 2018, Erlangen Ed. Andreas Maier, Thomas M. Deserno, Heinz Handels, Klaus H. Maier-Hein, Christoph Palm, Thomas Tolxdorff, Berlin, Heidelberg: Springer Vieweg, Berlin, Heidelberg, 2018. 341-346.

BibTeX: Download