An Adaptive Sampling Scheme to Efficiently Train Fully Convolutional Networks for Semantic Segmentation

Berger, Leslie; Hyde, Eoin; Cardoso, M. Jorge; Ourselin, Sébastien

doi:10.1007/978-3-319-95921-4_26

Cited by 28 publications

(17 citation statements)

References 25 publications

(34 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…To improve learning, an adaptive sampling method has been proposed (Berger et al, 2017) for DeepMedic. It consists in extracting more image patches in the regions where the prediction error is bigger, according to error maps generated throughout training.…”

Section: Methodsmentioning

confidence: 99%

Evaluation of Enhanced Learning Techniques for Segmenting Ischaemic Stroke Lesions in Brain Magnetic Resonance Perfusion Images Using a Convolutional Neural Network Scheme

Malla

Hernández

Rachmadi

et al. 2019

Front. Neuroinform.

View full text Add to dashboard Cite

Magnetic resonance (MR) perfusion imaging non-invasively measures cerebral perfusion, which describes the blood's passage through the brain's vascular network. Therefore, it is widely used to assess cerebral ischaemia. Convolutional Neural Networks (CNN) constitute the state-of-the-art method in automatic pattern recognition and hence, in segmentation tasks. But none of the CNN architectures developed to date have achieved high accuracy when segmenting ischaemic stroke lesions, being the main reasons their heterogeneity in location, shape, size, image intensity and texture, especially in this imaging modality. We use a freely available CNN framework, developed for MR imaging lesion segmentation, as core algorithm to evaluate the impact of enhanced machine learning techniques, namely data augmentation, transfer learning and post-processing, in the segmentation of stroke lesions using the ISLES 2017 dataset, which contains expert annotated diffusion-weighted perfusion and diffusion brain MRI of 43 stroke patients. Of all the techniques evaluated, data augmentation with binary closing achieved the best results, improving the mean Dice score in 17% over the baseline model. Consistent with previous works, better performance was obtained in the presence of large lesions.

show abstract

Section: Methodsmentioning

confidence: 99%

Evaluation of Enhanced Learning Techniques for Segmenting Ischaemic Stroke Lesions in Brain Magnetic Resonance Perfusion Images Using a Convolutional Neural Network Scheme

Malla

Hernández

Rachmadi

et al. 2019

Front. Neuroinform.

View full text Add to dashboard Cite

show abstract

Section: Methodsmentioning

confidence: 99%

Evaluation of Enhanced Learning Techniques for Segmenting Ischaemic Stroke Lesions in Brain Magnetic Resonance Perfusion Images using a Convolutional Neural Network Scheme

Malla

Valdés-Hernández

Rachmadi

et al. 2019

Preprint

View full text Add to dashboard Cite

2 3 Magnetic resonance (MR) perfusion imaging non-invasively measures cerebral perfusion, 4 which describes the blood's passage through the brain's vascular network. Therefore it is widely 5 used to assess cerebral ischaemia. Convolutional Neural Networks (CNN) constitute the state-6 of-the-art method in automatic pattern recognition and hence, in segmentation tasks. But none 7 of the CNN architectures developed to date have achieved high accuracy when segmenting 8 ischaemic stroke lesions, being the main reasons their heterogeneity in location, shape, size, 9 image intensity and texture, especially in this imaging modality. We use a freely available CNN 10 framework, developed for MR imaging lesion segmentation, as core algorithm to evaluate the 11 impact of enhanced machine learning techniques, namely data augmentation, transfer learning 12 and post-processing, in the segmentation of stroke lesions using the ISLES 2017 dataset, which 13 contains expert annotated diffusion-weighted perfusion and diffusion brain MRI of 43 stroke 14 patients. Of all the techniques evaluated, data augmentation with binary closing achieved the 15 best results, improving the mean Dice score in 17% over the baseline model. Consistent with 16 previous works, better performance was obtained in the presence of large lesions. 17

show abstract

“…They trained their model on simpler tasks before introducing the final problem of malignancy detection. Berger et al (2018) proposed to use an adaptive sampling strategy to improve the segmentation performance on difficult regions in multi-organ CT segmentation.…”

Section: Related Workmentioning

confidence: 99%

Automatic CNN-based detection of cardiac MR motion artefacts using k-space data augmentation and curriculum learning

Öksüz

Ruijsink

Puyol‐Antón

et al. 2019

Medical Image Analysis

View full text Add to dashboard Cite

Good quality of medical images is a prerequisite for the success of subsequent image analysis pipelines. Quality assessment of medical images is therefore an essential activity and for large population studies such as the UK Biobank (UKBB), manual identification of artefacts such as those caused by unanticipated motion is tedious and time-consuming. Therefore, there is an urgent need for automatic image quality assessment techniques. In this paper, we propose a method to automatically detect the presence of motion-related artefacts in cardiac magnetic resonance (CMR) cine images. We compare two deep learning architectures to classify poor quality CMR images: 1) 3D spatio-temporal Convolutional Neural Networks (3D-CNN), 2) Long-term Recurrent Convolutional Network (LRCN). Though in real clinical setup motion artefacts are common, high-quality imaging of UKBB, which comprises cross-sectional population data of volunteers who do not necessarily have health problems creates a highly imbalanced classification problem. Due to the high number of good quality images compared to the relatively low number of images with motion artefacts, we propose a novel data augmentation scheme based on synthetic artefact creation in k-space. We also investigate a learning approach using a predetermined curriculum based on synthetic artefact severity. We evaluate our pipeline on a subset of the UK Biobank data set consisting of 3510 CMR images. The LRCN architecture outperformed the 3D-CNN architecture and was able to detect 2D+time short axis images with motion artefacts in less than 1ms with high recall. We compare our approach to a range of state-of-the-art quality assessment methods. The novel data augmentation and curriculum learning approaches both improved classification performance achieving overall area under the ROC curve of 0.89.

show abstract

An Adaptive Sampling Scheme to Efficiently Train Fully Convolutional Networks for Semantic Segmentation

Cited by 28 publications

References 25 publications

Evaluation of Enhanced Learning Techniques for Segmenting Ischaemic Stroke Lesions in Brain Magnetic Resonance Perfusion Images Using a Convolutional Neural Network Scheme

Evaluation of Enhanced Learning Techniques for Segmenting Ischaemic Stroke Lesions in Brain Magnetic Resonance Perfusion Images Using a Convolutional Neural Network Scheme

Evaluation of Enhanced Learning Techniques for Segmenting Ischaemic Stroke Lesions in Brain Magnetic Resonance Perfusion Images using a Convolutional Neural Network Scheme

Automatic CNN-based detection of cardiac MR motion artefacts using k-space data augmentation and curriculum learning

Contact Info

Product

Resources

About