Open AccessAn automatic method for prostate segmentation on 3D MRI scans using local phylogenetic indexes and XGBoost
Author(s) -
Giovanni L. F. da Silva,
Francisco Y. C. de Oliveira,
João Otávio Bandeira Diniz,
Petterson Sousa Diniz,
Darlan B. P. Quintanilha,
Aristófanes Corrêa Silva,
Anselmo Cardoso de Paiva,
Elton Anderson Araújo de Cavalcanti
Publication year - 2021
Language(s) - English
Resource type - Conference proceedings
DOI - 10.5753/sbcas.2021.16062
Subject(s) - segmentation , dice , prostate , similarity (geometry) , prostate cancer , computer science , artificial intelligence , magnetic resonance imaging , t2 weighted , image segmentation , pattern recognition (psychology) , computer vision , medicine , radiology , mathematics , image (mathematics) , cancer , statistics
The detection, diagnosis, and treatment of prostate cancer depends on the correct determination of the prostate anatomy. In current practice, the prostate segmentation is performed manually by a radiologist, which is extremely time-consuming that demands experience and concentration. Therefore, this paper proposes an automatic method for prostate segmentation on 3D magnetic resonance imaging scans using a superpixel technique, phylogenetic indexes, and an optimized XGBoost algorithm. The proposed method has been evaluated on the Prostate 3T and PROMISE12 databases presenting a dice similarity coefcient of 84.48% and a volumetric similarity of 95.91%, demonstrating the high-performance potential of the proposed method.