ObjectiveThe aim of this paper is to investigate the use of fully convolutional neural networks (FCNNs) to segment scar tissue in the left ventricle from cardiac magnetic resonance with late gadolinium enhancement (CMR-LGE) images.MethodsA successful FCNN in the literature (the ENet) was modified and trained to provide scar-tissue segmentation. Two segmentation protocols (Protocol 1 and Protocol 2) were investigated, the latter limiting the scar-segmentation search area to the left ventricular myocardial tissue region. CMR-LGE from 30 patients with ischemic-heart disease were retrospectively analyzed, for a total of 250 images, presenting high variability in terms of scar dimension and location. Segmentation results were assessed against manual scar-tissue tracing using one-patient-out cross validation.ResultsProtocol 2 outperformed Protocol 1 significantly (p value < 0.05), with median sensitivity and Dice similarity coefficient equal to 88.07% [inter-quartile range (IQR) 18.84%] and 71.25% (IQR 31.82%), respectively.DiscussionBoth segmentation protocols were able to detect scar tissues in the CMR-LGE images but higher performance was achieved when limiting the search area to the myocardial region. The findings of this paper represent an encouraging starting point for the use of FCNNs for the segmentation of nonviable scar tissue from CMR-LGE images.

Development and testing of a deep learning-based strategy for scar segmentation on CMR-LGE images / S. Moccia, R. Banali, C. Martini, G. Muscogiuri, G. Pontone, M. Pepi, E.G. Caiani. - In: MAGNETIC RESONANCE MATERIALS IN PHYSICS BIOLOGY AND MEDICINE. - ISSN 0968-5243. - 32:2(2019 Apr), pp. 187-195. [10.1007/s10334-018-0718-4]

Development and testing of a deep learning-based strategy for scar segmentation on CMR-LGE images

G. Pontone;
2019

Abstract

ObjectiveThe aim of this paper is to investigate the use of fully convolutional neural networks (FCNNs) to segment scar tissue in the left ventricle from cardiac magnetic resonance with late gadolinium enhancement (CMR-LGE) images.MethodsA successful FCNN in the literature (the ENet) was modified and trained to provide scar-tissue segmentation. Two segmentation protocols (Protocol 1 and Protocol 2) were investigated, the latter limiting the scar-segmentation search area to the left ventricular myocardial tissue region. CMR-LGE from 30 patients with ischemic-heart disease were retrospectively analyzed, for a total of 250 images, presenting high variability in terms of scar dimension and location. Segmentation results were assessed against manual scar-tissue tracing using one-patient-out cross validation.ResultsProtocol 2 outperformed Protocol 1 significantly (p value < 0.05), with median sensitivity and Dice similarity coefficient equal to 88.07% [inter-quartile range (IQR) 18.84%] and 71.25% (IQR 31.82%), respectively.DiscussionBoth segmentation protocols were able to detect scar tissues in the CMR-LGE images but higher performance was achieved when limiting the search area to the myocardial region. The findings of this paper represent an encouraging starting point for the use of FCNNs for the segmentation of nonviable scar tissue from CMR-LGE images.
CMR-LGE images; Deep learning; Fully-convolutional neural networks; Scar segmentation
Settore MED/11 - Malattie dell'Apparato Cardiovascolare
   Linking excellence in biomedical knowledge and computational intelligence research for personalized management of CVD within PHC
   LINK
   European Commission
   Horizon 2020 Framework Programme
   692023
apr-2019
Article (author)
File in questo prodotto:
File Dimensione Formato  
development.pdf

accesso riservato

Descrizione: Research Article
Tipologia: Publisher's version/PDF
Dimensione 1.15 MB
Formato Adobe PDF
1.15 MB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2434/955912
Citazioni
  • ???jsp.display-item.citation.pmc??? 26
  • Scopus 83
  • ???jsp.display-item.citation.isi??? 69
social impact