@inproceedings{0163914035254135b37bbb3a4d0197f1,
title = "Automatic analysis of pediatric renal ultrasound using shape, anatomical and image acquisition priors",
abstract = "In this paper we present a segmentation method for ultrasound (US) images of the pediatric kidney, a difficult and barely studied problem. Our method segments the kidney on 2D sagittal US images and relies on minimal user intervention and a combination of improvements made to the Active Shape Model (ASM) framework. Our contributions include particle swarm initialization and profile training with rotation correction. We also introduce our methodology for segmentation of the kidney's collecting system (CS), based on graph-cuts (GC) with intensity and positional priors. Our intensity model corrects for intensity bias by comparison with other biased versions of the most similar kidneys in the training set. We prove significant improvements (p < 0.001) with respect to classic ASM and GC for kidney and CS segmentation, respectively. We use our semi-automatic method to compute the hydronephrosis index (HI) with an average error of 2.67±5.22 percentage points similar to the error of manual HI between different operators of 2.31±4.54 percentage points.",
author = "Mendoza, {Carlos S.} and Xin Kang and Nabile Safdar and Emmarie Myers and Martin, {Aaron D.} and Enrico Grisan and Peters, {Craig A.} and Linguraru, {Marius George}",
year = "2013",
doi = "10.1007/978-3-642-40760-4_33",
language = "English",
isbn = "9783642407598",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Nature",
number = "PART 3",
pages = "259--266",
booktitle = "Medical Image Computing and Computer-Assisted Intervention, MICCAI 2013 - 16th International Conference, Proceedings",
edition = "PART 3",
note = "16th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2013 ; Conference date: 22-09-2013 Through 26-09-2013",
}