Efficient and Motion Correction-Free Myocardial Perfusion Segmentation in Small MRI Data Using Deep Transfer Learning From Cine Images: A Promising Framework for Clinical Implementation

Perfusion cardiovascular magnetic resonance imaging is used to quantify the heart’s blood flow, which requires the segmentation of the myocardium, a laborious task. Deep learning-based methods, the most accurate to accomplish this task, still rely on expensive motion correction steps and...

Full description

Saved in:
Bibliographic Details
Main Authors: German Garcia-Jara, Angel Jimenez-Molina, Esteban Reyes, Nicolas Tapia-Rivas, Cristobal Ramos-Gomez, Jose De Grazia, Matias Sepulveda
Format: Article
Language:English
Published: IEEE 2023-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10247052/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849727981163905024
author German Garcia-Jara
Angel Jimenez-Molina
Esteban Reyes
Nicolas Tapia-Rivas
Cristobal Ramos-Gomez
Jose De Grazia
Matias Sepulveda
author_facet German Garcia-Jara
Angel Jimenez-Molina
Esteban Reyes
Nicolas Tapia-Rivas
Cristobal Ramos-Gomez
Jose De Grazia
Matias Sepulveda
author_sort German Garcia-Jara
collection DOAJ
description Perfusion cardiovascular magnetic resonance imaging is used to quantify the heart’s blood flow, which requires the segmentation of the myocardium, a laborious task. Deep learning-based methods, the most accurate to accomplish this task, still rely on expensive motion correction steps and require large labeled datasets. This paper presents an innovative, efficient approach to myocardial perfusion segmentation, utilizing deep learning techniques without motion correction and with minimal data requirements. Through transfer learning, this methodology leverages the wealth of information available from large, publicly accessible cine magnetic resonance datasets, which provide anatomically analogous images to perfusion ones. This methodology includes normalization and cropping of cine images using a Region-of-Interest detector based on a Markovian, graph-based visual saliency algorithm improved by a sequence of morphological operations. After pretraining a U-net convolutional neural network, a special fine-tuning scheme optimizes its performance. The parameters learned are the starting point for training on a smaller perfusion dataset from the Clinical Hospital of the University of Chile. After an ablation study, the best model is obtained when using both cropping and fine-tuning from the cine dataset, segmenting the left ventricle endocardium with Dice, IoU, and Hausdorff distance of 92.2%, 85.9%, and 5.1 mm respectively, and 95.6%, 91.7%, and 4.6 mm for the left ventricle epicardium. Notably, fine-tuning achieves a Dice of 91.8% for endocardium and 95.2% for epicardium when only 289 perfusion training images are available. These are promising results for developing targeted implementations in real healthcare settings when only small datasets are available.
format Article
id doaj-art-83d4e47d42584d43998ba649dc7ef330
institution DOAJ
issn 2169-3536
language English
publishDate 2023-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-83d4e47d42584d43998ba649dc7ef3302025-08-20T03:09:42ZengIEEEIEEE Access2169-35362023-01-011110317710318810.1109/ACCESS.2023.331398010247052Efficient and Motion Correction-Free Myocardial Perfusion Segmentation in Small MRI Data Using Deep Transfer Learning From Cine Images: A Promising Framework for Clinical ImplementationGerman Garcia-Jara0https://orcid.org/0000-0001-8202-9314Angel Jimenez-Molina1https://orcid.org/0000-0002-6866-6584Esteban Reyes2Nicolas Tapia-Rivas3https://orcid.org/0000-0002-6501-4687Cristobal Ramos-Gomez4Jose De Grazia5Matias Sepulveda6https://orcid.org/0009-0005-5384-8296Department of Electrical Engineering, Faculty of Physical and Mathematical Sciences, University of Chile, Santiago, ChileDepartment of Industrial Engineering, Faculty of Physical and Mathematical Sciences, University of Chile, Santiago, ChileDepartment of Electrical Engineering, Faculty of Physical and Mathematical Sciences, University of Chile, Santiago, ChileDepartment of Electrical Engineering, Faculty of Physical and Mathematical Sciences, University of Chile, Santiago, ChileDepartment of Radiology, Faculty of Medicine, University of Chile, Santiago, ChileDepartment of Radiology, Faculty of Medicine, University of Chile, Santiago, ChileDepartment of Industrial Engineering, Faculty of Physical and Mathematical Sciences, University of Chile, Santiago, ChilePerfusion cardiovascular magnetic resonance imaging is used to quantify the heart’s blood flow, which requires the segmentation of the myocardium, a laborious task. Deep learning-based methods, the most accurate to accomplish this task, still rely on expensive motion correction steps and require large labeled datasets. This paper presents an innovative, efficient approach to myocardial perfusion segmentation, utilizing deep learning techniques without motion correction and with minimal data requirements. Through transfer learning, this methodology leverages the wealth of information available from large, publicly accessible cine magnetic resonance datasets, which provide anatomically analogous images to perfusion ones. This methodology includes normalization and cropping of cine images using a Region-of-Interest detector based on a Markovian, graph-based visual saliency algorithm improved by a sequence of morphological operations. After pretraining a U-net convolutional neural network, a special fine-tuning scheme optimizes its performance. The parameters learned are the starting point for training on a smaller perfusion dataset from the Clinical Hospital of the University of Chile. After an ablation study, the best model is obtained when using both cropping and fine-tuning from the cine dataset, segmenting the left ventricle endocardium with Dice, IoU, and Hausdorff distance of 92.2%, 85.9%, and 5.1 mm respectively, and 95.6%, 91.7%, and 4.6 mm for the left ventricle epicardium. Notably, fine-tuning achieves a Dice of 91.8% for endocardium and 95.2% for epicardium when only 289 perfusion training images are available. These are promising results for developing targeted implementations in real healthcare settings when only small datasets are available.https://ieeexplore.ieee.org/document/10247052/Deep transfer learningfine-tuningmyocardial perfusionmyocardial segmentation frameworkU-net convolutional neural network
spellingShingle German Garcia-Jara
Angel Jimenez-Molina
Esteban Reyes
Nicolas Tapia-Rivas
Cristobal Ramos-Gomez
Jose De Grazia
Matias Sepulveda
Efficient and Motion Correction-Free Myocardial Perfusion Segmentation in Small MRI Data Using Deep Transfer Learning From Cine Images: A Promising Framework for Clinical Implementation
IEEE Access
Deep transfer learning
fine-tuning
myocardial perfusion
myocardial segmentation framework
U-net convolutional neural network
title Efficient and Motion Correction-Free Myocardial Perfusion Segmentation in Small MRI Data Using Deep Transfer Learning From Cine Images: A Promising Framework for Clinical Implementation
title_full Efficient and Motion Correction-Free Myocardial Perfusion Segmentation in Small MRI Data Using Deep Transfer Learning From Cine Images: A Promising Framework for Clinical Implementation
title_fullStr Efficient and Motion Correction-Free Myocardial Perfusion Segmentation in Small MRI Data Using Deep Transfer Learning From Cine Images: A Promising Framework for Clinical Implementation
title_full_unstemmed Efficient and Motion Correction-Free Myocardial Perfusion Segmentation in Small MRI Data Using Deep Transfer Learning From Cine Images: A Promising Framework for Clinical Implementation
title_short Efficient and Motion Correction-Free Myocardial Perfusion Segmentation in Small MRI Data Using Deep Transfer Learning From Cine Images: A Promising Framework for Clinical Implementation
title_sort efficient and motion correction free myocardial perfusion segmentation in small mri data using deep transfer learning from cine images a promising framework for clinical implementation
topic Deep transfer learning
fine-tuning
myocardial perfusion
myocardial segmentation framework
U-net convolutional neural network
url https://ieeexplore.ieee.org/document/10247052/
work_keys_str_mv AT germangarciajara efficientandmotioncorrectionfreemyocardialperfusionsegmentationinsmallmridatausingdeeptransferlearningfromcineimagesapromisingframeworkforclinicalimplementation
AT angeljimenezmolina efficientandmotioncorrectionfreemyocardialperfusionsegmentationinsmallmridatausingdeeptransferlearningfromcineimagesapromisingframeworkforclinicalimplementation
AT estebanreyes efficientandmotioncorrectionfreemyocardialperfusionsegmentationinsmallmridatausingdeeptransferlearningfromcineimagesapromisingframeworkforclinicalimplementation
AT nicolastapiarivas efficientandmotioncorrectionfreemyocardialperfusionsegmentationinsmallmridatausingdeeptransferlearningfromcineimagesapromisingframeworkforclinicalimplementation
AT cristobalramosgomez efficientandmotioncorrectionfreemyocardialperfusionsegmentationinsmallmridatausingdeeptransferlearningfromcineimagesapromisingframeworkforclinicalimplementation
AT josedegrazia efficientandmotioncorrectionfreemyocardialperfusionsegmentationinsmallmridatausingdeeptransferlearningfromcineimagesapromisingframeworkforclinicalimplementation
AT matiassepulveda efficientandmotioncorrectionfreemyocardialperfusionsegmentationinsmallmridatausingdeeptransferlearningfromcineimagesapromisingframeworkforclinicalimplementation