Rotation invariance and directional sensitivity : spherical harmonics versus radiomics features

Depeursinge, Adrien (Biomedical Imaging Group, Ecole Polytechnique Fédérale de Lausanne (EPFL) ; University of Applied Sciences and Arts Western Switzerland (HES-SO Valais-Wallis)) ; Fageot, Julien (Biomedical Imaging Group, Ecole Polytechnique F_ed_erale de Lausanne (EPFL)) ; Andrearczyk, Vincent (University of Applied Sciences and Arts Western Switzerland (HES-SO Valais-Wallis)) ; Ward, John Paul (Department of Mathematics, North Carolina A&T State University, NC, USA) ; Unser, Michael (Biomedical Imaging Group, Ecole Polytechnique Fédérale de Lausanne (EPFL))

We define and investigate the Local Rotation Invariance (LRI) and Directional Sensitivity (DS) of radiomics features. Most of the classical features cannot combine the two properties, which are antagonist in simple designs. We propose texture operators based on spherical harmonic wavelets (SHW) invariants and show that they are both LRI and DS. An experimental comparison of SHW and popular radiomics operators for classifying 3D textures reveals the importance of combining the two properties for optimal pattern characterization.


Keywords:
Faculty:
Economie et Services
School:
HEG-VS
Institute:
Institut Informatique de gestion
Subject(s):
Informatique
Publisher:
Cham, Springer
Date:
2018-09
Cham
Springer
Pagination:
pp. 107-115
Published in:
Machine learning in medical imaging
Author of the book:
Shi, Yinghuan ; ed. ; Nanjing University, China
Suk, Heung-Il ; ed. ; Korea University, Seoul
Liu, Minigxia ; ed. ; Liu, Minigxia
DOI:
ISBN:
978-3-030-00918-2
External resources:
Appears in Collection:

Note: The status of this file is: restricted


 Record created 2018-11-02, last modified 2018-12-20

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)