Improving cardiac MRI convolutional neural network segmentation on small training datasets and dataset shift: A continuous kernel cut approach

Guo, Fumin; Ng, Matthew; Goubran, Maged; Petersen, Steffen E.; Piechnik, Stefan K.; Neubauer, Stefan; Wright, Graham A.

doi:10.1016/j.media.2020.101636

Cited by 48 publications

(46 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Results indicate that both, size and diversity of the training data, are relevant. State-of-the-art results can be achieved with images from only 36 patients which is in line with previous works [41] achieving a mean DSC of 0.99 on LTRC test data using the U-net(R-36) model.…”

Section: Discussionsupporting

confidence: 88%

Automatic lung segmentation in routine imaging is primarily a data diversity problem, not a methodology problem

et al. 2020

View full text Add to dashboard Cite

Background: Automated segmentation of anatomical structures is a crucial step in image analysis. For lung segmentation in computed tomography, a variety of approaches exists, involving sophisticated pipelines trained and validated on different datasets. However, the clinical applicability of these approaches across diseases remains limited. Methods: We compared four generic deep learning approaches trained on various datasets and two readily available lung segmentation algorithms. We performed evaluation on routine imaging data with more than six different disease patterns and three published data sets. Results: Using different deep learning approaches, mean Dice similarity coefficients (DSCs) on test datasets varied not over 0.02. When trained on a diverse routine dataset (n = 36), a standard approach (U-net) yields a higher DSC (0.97 ± 0.05) compared to training on public datasets such as the Lung Tissue Research Consortium (0.94 ± 0.13, p = 0.024) or Anatomy 3 (0.92 ± 0.15, p = 0.001). Trained on routine data (n = 231) covering multiple diseases, U-net compared to reference methods yields a DSC of 0.98 ± 0.03 versus 0.94 ± 0.12 (p = 0.024). Conclusions: The accuracy and reliability of lung segmentation algorithms on demanding cases primarily relies on the diversity of the training data, highlighting the importance of data diversity compared to model choice. Efforts in developing new datasets and providing trained models to the public are critical. By releasing the trained model under General Public License 3.0, we aim to foster research on lung diseases by providing a readily available tool for segmentation of pathological lungs.

show abstract

Section: Discussionsupporting

confidence: 88%

Automatic lung segmentation in routine imaging is primarily a data diversity problem, not a methodology problem

et al. 2020

View full text Add to dashboard Cite

show abstract

“…In particular, cine myocardium segmentation and cine‐LGE registration were evaluated both globally for the entire myocardium and regionally for the region containing scar, as illustrated in Supporting Information Figure . Cine myocardium segmentation accuracies were determined using dice similarity coefficient (DSC) and average symmetric surface‐distance (ASSD) 32 between the U‐net and manual cine myocardium masks. Cine‐LGE registration accuracy was quantified by comparing the warped cine manual masks and the LGE myocardium manual segmentation using DSC and ASSD to only evaluate the effects of the registrations, as previously suggested 33,34 .…”

Section: Methodsmentioning

confidence: 99%

Cine and late gadolinium enhancement MRI registration and automated myocardial infarct heterogeneity quantification

Guo

Krahn

Roifman

et al. 2020

Magnetic Resonance in Med

Self Cite

View full text Add to dashboard Cite

Purpose To develop an approach for automated quantification of myocardial infarct heterogeneity in late gadolinium enhancement (LGE) cardiac MRI. Methods We acquired 2D short‐axis cine and 3D LGE in 10 pigs with myocardial infarct. The 2D cine myocardium was segmented and registered to the LGE images. LGE image signal intensities within the warped cine myocardium masks were analyzed to determine the thresholds of infarct core (IC) and gray zone (GZ) for the standard‐deviation (SD) and full‐width‐at‐halfmaximum (FWHM) methods. The initial IC, GZ, and IC + GZ segmentations were postprocessed using a normalized cut approach. Cine segmentation and cine‐LGE registration accuracies were evaluated using dice similarity coefficient and average symmetric surface distance. Automated IC, GZ, and IC + GZ volumes were compared with manual results using Pearson correlation coefficient (r), Bland‐Altman analyses, and intraclass correlation coefficient. Results For n = 87 slices containing scar, we achieved cine segmentation dice similarity coefficient = 0.87 ± 0.12, average symmetric surface distance = 0.94 ± 0.74 mm (epicardium), and 1.03 ± 0.82 mm (endocardium) in the scar region. For cine‐LGE registration, dice similarity coefficient was 0.90 ± 0.06 and average symmetric surface distance was 0.72 ± 0.39 mm (epicardium) and 0.86 ± 0.53 mm (endocardium) in the scar region. For both SD and FWHM methods, automated IC, GZ, and IC + GZ volumes were strongly (r > 0.70) correlated with manual measurements, and the correlations were not significantly different from interobserver correlations (P > .05). The agreement between automated and manual scar volumes (intraclass correlation coefficient = 0.85‐0.96) was similar to that between two observers (intraclass correlation coefficient = 0.81‐0.99); automated scar segmentation errors were not significantly different from interobserver segmentation differences (P > .05). Conclusions Our approach provides fully automated cine‐LGE MRI registration and LGE myocardial infarct heterogeneity quantification in preclinical studies.

show abstract

“…MRI of sub-cortical brain structure automatic and accurate segmentation using CNN to extract prior spatial features and train the methods on most of complicated features to improve accuracy which is effective for the processes, such as pre-operative evaluation, surgical planning, radiotherapy treatment planning, and longitudinal monitoring for disease progression [20]. It provides a wealth of imaging biomarkers for cardiovascular disease care and segmentation of cardiac structures [21]. Furthermore, it provides rich information about the human tissue anatomies so as to earn soft-tissue contrast widely.…”

Section: Medical Image Modalitiesmentioning

confidence: 99%

Review Research of Medical Image Analysis Using Deep Learning

Mohammed

Al-Ani

2020

UHD J SCI TECH

View full text Add to dashboard Cite

In modern globe, medical image analysis significantly participates in diagnosis process. In general, it involves five processes, such as medical image classification, medical image detection, medical image segmentation, medical image registration, and medical image localization. Medical imaging uses in diagnosis process for most of the human body organs, such as brain tumor, chest, breast, colonoscopy, retinal, and many other cases relate to medical image analysis using various modalities. Multi-modality images include magnetic resonance imaging, single photon emission computed tomography (CT), positron emission tomography, optical coherence tomography, confocal laser endoscopy, magnetic resonance spectroscopy, CT, X-ray, wireless capsule endoscopy, breast cancer, papanicolaou smear, hyper spectral image, and ultrasound use to diagnose different body organs and cases. Medical image analysis is appropriate environment to interact with automate intelligent system technologies. Among the intelligent systems deep learning (DL) is the modern one to manipulate medical image analysis processes and processing an image into fundamental components to extract meaningful information. The best model to establish its systems is deep convolutional neural network. This study relied on reviewing of some of these studies because of these reasons; improvements of medical imaging increase demand on automate systems of medical image analysis using DL, in most tested cases, accuracy of intelligent methods especially DL methods higher than accuracy of hand-crafted works. Furthermore, manually works need a lot of time compare to systematic diagnosis.

show abstract

Improving cardiac MRI convolutional neural network segmentation on small training datasets and dataset shift: A continuous kernel cut approach

Cited by 48 publications

References 47 publications

Automatic lung segmentation in routine imaging is primarily a data diversity problem, not a methodology problem

Automatic lung segmentation in routine imaging is primarily a data diversity problem, not a methodology problem

Cine and late gadolinium enhancement MRI registration and automated myocardial infarct heterogeneity quantification

Review Research of Medical Image Analysis Using Deep Learning

Contact Info

Product

Resources

About