Topology-preserving augmentation for cnn-based segmentation of congenital heart defects from 3d paediatric cmr

Nick Byrne*, James R. Clough, Isra Valverde, Giovanni Montana, Andrew P. King

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference paperpeer-review

2 Citations (Scopus)
70 Downloads (Pure)

Abstract

Patient-specific 3D printing of congenital heart anatomy demands an accurate segmentation of the thin tissue interfaces which characterise these diagnoses. Even when a label set has a high spatial overlap with the ground truth, inaccurate delineation of these interfaces can result in topological errors. These compromise the clinical utility of such models due to the anomalous appearance of defects. CNNs have achieved state-of-the-art performance in segmentation tasks. Whilst data augmentation has often played an important role, we show that conventional image resampling schemes used therein can introduce topological changes in the ground truth labelling of augmented samples. We present a novel pipeline to correct for these changes, using a fast-marching algorithm to enforce the topology of the ground truth labels within their augmented representations. In so doing, we invoke the idea of cardiac contiguous topology to describe an arbitrary combination of congenital heart defects and develop an associated, clinically meaningful metric to measure the topological correctness of segmentations. In a series of five-fold cross-validations, we demonstrate the performance gain produced by this pipeline and the relevance of topological considerations to the segmentation of congenital heart defects. We speculate as to the applicability of this approach to any segmentation task involving morphologically complex targets.

Original languageEnglish
Title of host publicationSmart Ultrasound Imaging and Perinatal, Preterm and Paediatric Image Analysis - 1st International Workshop, SUSI 2019, and 4th International Workshop, PIPPI 2019, Held in Conjunction with MICCAI 2019, Proceedings
EditorsQian Wang, Alberto Gomez, Jana Hutter, Alberto Gomez, Veronika Zimmer, Jana Hutter, Emma Robinson, Daan Christiaens, Andrew Melbourne, Kristin McLeod, Oliver Zettinig, Roxane Licandro, Esra Abaci Turk
PublisherSPRINGER
Pages181-188
Number of pages8
ISBN (Print)9783030328740
DOIs
Publication statusPublished - 2019
Event1st International Workshop on Smart Ultrasound Imaging, SUSI 2019, and the 4th International Workshop on Preterm, Perinatal and Paediatric Image Analysis, PIPPI 2019, held in conjunction with the 22nd International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2019 - Shenzhen, China
Duration: 17 Oct 201917 Oct 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11798 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference1st International Workshop on Smart Ultrasound Imaging, SUSI 2019, and the 4th International Workshop on Preterm, Perinatal and Paediatric Image Analysis, PIPPI 2019, held in conjunction with the 22nd International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2019
Country/TerritoryChina
CityShenzhen
Period17/10/201917/10/2019

Keywords

  • Data augmentation
  • Image segmentation
  • Topology

Fingerprint

Dive into the research topics of 'Topology-preserving augmentation for cnn-based segmentation of congenital heart defects from 3d paediatric cmr'. Together they form a unique fingerprint.

Cite this