OBELISK-Net: Fewer layers to solve 3D multi-organ segmentation with sparse deformable convolutions

Heinrich, Mattias P.; Oktay, Ozan; Bouteldja, Nassim

doi:10.1016/j.media.2019.02.006

Cited by 67 publications

(42 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…1) Convolutional feature learning network: To learn a meaningful nonlinear mapping from input intensities to a dense feature volume (with |c| = 16 channels and a stride of 3), we employ the Obelisk approach [4], which comprises a 3D deformable convolution with trainable offsets followed by a simple 1 × 1 MLP and captures spatial context very effectively. We extend the authors' implementation by adding a normal 5 × 5 × 5 convolution kernel with 4 channels prior to the Obelisk layer to also learn edge-like features.…”

Section: Methodsmentioning

confidence: 99%

Closing the Gap Between Deep and Conventional Image Registration Using Probabilistic Dense Displacement Networks

Heinrich

2019

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Nonlinear image registration continues to be a fundamentally important tool in medical image analysis. Diagnostic tasks, imageguided surgery and radiotherapy as well as motion analysis all rely heavily on accurate intra-patient alignment. Furthermore, inter-patient registration enables atlas-based segmentation or landmark localisation and shape analysis. When labelled scans are scarce and anatomical differences large, conventional registration has often remained superior to deep learning methods that have so far mainly dealt with relatively small or low-complexity deformations. We address this shortcoming by leveraging ideas from probabilistic dense displacement optimisation that has excelled in many registration tasks with large deformations. We propose to design a network with approximate min-convolutions and mean field inference for differentiable displacement regularisation within a discrete weakly-supervised registration setting. By employing these meaningful and theoretically proven constraints, our learnable registration algorithm contains very few trainable weights (primarily for feature extraction) and is easier to train with few labelled scans. It is very fast in training and inference and achieves state-of-the-art accuracies for the challenging inter-patient registration of abdominal CT outperforming previous deep learning approaches by 15% Dice overlap.

show abstract

Section: Methodsmentioning

confidence: 99%

Closing the Gap Between Deep and Conventional Image Registration Using Probabilistic Dense Displacement Networks

Heinrich

2019

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…The authors claimed highly precise results for segmenting cervical tumors on the 3D PET. In Reference [ 73 ], the authors proposed 3D convolution kernels for learning filter coefficients and spatial filter offsets simultaneously for 3D CT multi-organ segmentation work. The outcomes were compared to U-Net architectures and the authors claim that their architecture requires less trainable parameters and storage to obtain a high quality.…”

Section: Applications In 3d Medical Imagingmentioning

confidence: 99%

3D Deep Learning on Medical Images: A Review

Singh

Wang

Gupta

et al. 2020

Sensors

344

150

View full text Add to dashboard Cite

The rapid advancements in machine learning, graphics processing technologies and the availability of medical imaging data have led to a rapid increase in the use of deep learning models in the medical domain. This was exacerbated by the rapid advancements in convolutional neural network (CNN) based architectures, which were adopted by the medical imaging community to assist clinicians in disease diagnosis. Since the grand success of AlexNet in 2012, CNNs have been increasingly used in medical image analysis to improve the efficiency of human clinicians. In recent years, three-dimensional (3D) CNNs have been employed for the analysis of medical images. In this paper, we trace the history of how the 3D CNN was developed from its machine learning roots, we provide a brief mathematical description of 3D CNN and provide the preprocessing steps required for medical images before feeding them to 3D CNNs. We review the significant research in the field of 3D medical imaging analysis using 3D CNNs (and its variants) in different medical areas such as classification, segmentation, detection and localization. We conclude by discussing the challenges associated with the use of 3D CNNs in the medical imaging domain (and the use of deep learning models in general) and possible future trends in the field.

show abstract

“…The novel idea of DeepLab v3 is replacing the last ResNet block by an Atrous Spatial Pyramid Pooling (ASPP) block. The third architecture is the Obelisk-Net by Heinrich et al [13], which employs sparse convolutions with arbitrary offsets to the kernel center. These sparse convolutions have the advantage of drastically decreasing the number of parameters in the model while keeping a large receptive field.…”

Section: Approachmentioning

confidence: 99%

DL-based segmentation of endoscopic scenes for mitral valve repair

Ivantsits

Tautz

Sündermann

et al. 2020

Current Directions in Biomedical Engineering

View full text Add to dashboard Cite

Minimally invasive surgery is increasingly utilized for mitral valve repair and replacement. The intervention is performed with an endoscopic field of view on the arrested heart. Extracting the necessary information from the live endoscopic video stream is challenging due to the moving camera position, the high variability of defects, and occlusion of structures by instruments. During such minimally invasive interventions there is no time to segment regions of interest manually. We propose a real-time-capable deep-learning-based approach to detect and segment the relevant anatomical structures and instruments. For the universal deployment of the proposed solution, we evaluate them on pixel accuracy as well as distance measurements of the detected contours. The U-Net, Google’s DeepLab v3, and the Obelisk-Net models are cross-validated, with DeepLab showing superior results in pixel accuracy and distance measurements.

show abstract

OBELISK-Net: Fewer layers to solve 3D multi-organ segmentation with sparse deformable convolutions

Cited by 67 publications

References 13 publications

Closing the Gap Between Deep and Conventional Image Registration Using Probabilistic Dense Displacement Networks

Closing the Gap Between Deep and Conventional Image Registration Using Probabilistic Dense Displacement Networks

3D Deep Learning on Medical Images: A Review

DL-based segmentation of endoscopic scenes for mitral valve repair

Contact Info

Product

Resources

About