Image Segmentation, Registration and Characterization in <i>R</i> with <b>SimpleITK</b>

Beare, Richard; Lowekamp, Bradley; Yaniv, Ziv

doi:10.18637/jss.v086.i08

Cited by 152 publications

(95 citation statements)

References 40 publications

(38 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The pre-processing tasks employed in brain image processing include image registration, zero padding, and histogram matching-based intensity normalization. The first step is a rigid registration from the target image to the MNI305 template (256 × 256 × 256) using 3D rigid registration with an Euler transform (SimpleITK) [ 49 , 50 , 51 , 52 ]. Next, zero padding is performed: 16 × 16 × 16 padding is used for training and 24 × 24 × 24 is used for testing.…”

Section: Methodsmentioning

confidence: 99%

Split-Attention U-Net: A Fully Convolutional Network for Robust Multi-Label Segmentation from Brain MRI

Kim

et al. 2020

Brain Sciences

View full text Add to dashboard Cite

Multi-label brain segmentation from brain magnetic resonance imaging (MRI) provides valuable structural information for most neurological analyses. Due to the complexity of the brain segmentation algorithm, it could delay the delivery of neuroimaging findings. Therefore, we introduce Split-Attention U-Net (SAU-Net), a convolutional neural network with skip pathways and a split-attention module that segments brain MRI scans. The proposed architecture employs split-attention blocks, skip pathways with pyramid levels, and evolving normalization layers. For efficient training, we performed pre-training and fine-tuning with the original and manually modified FreeSurfer labels, respectively. This learning strategy enables involvement of heterogeneous neuroimaging data in the training without the need for many manual annotations. Using nine evaluation datasets, we demonstrated that SAU-Net achieved better segmentation accuracy with better reliability that surpasses those of state-of-the-art methods. We believe that SAU-Net has excellent potential due to its robustness to neuroanatomical variability that would enable almost instantaneous access to accurate neuroimaging biomarkers and its swift processing runtime compared to other methods investigated.

show abstract

Section: Methodsmentioning

confidence: 99%

Split-Attention U-Net: A Fully Convolutional Network for Robust Multi-Label Segmentation from Brain MRI

Kim

et al. 2020

Brain Sciences

View full text Add to dashboard Cite

show abstract

“…These plugins do not necessarily need to be GPU-accelerated. To demonstrate the extensibility, about 50 additional image processing operations, utilizing the open-source libraries ImageJ, ImageJ2, Imglib2 [10], BoneJ [11], MorpholibJ [12], the ImageJ 3D Suite [13] and SimpleITK [14], are available for installation and testing via a separate Fiji update site.…”

Section: Methodsmentioning

confidence: 99%

Interactive design of GPU-accelerated Image Data Flow Graphs and cross-platform deployment using multi-lingual code generation

Haase

Jain

Rigaud

et al. 2020

Preprint

View full text Add to dashboard Cite

Modern life science relies heavily on fluorescent microscopy and subsequent quantitative bio-image analysis. The current rise of graphics processing units (GPUs) in the context of image processing enables batch processing large amounts of image data at unprecedented speed. In order to facilitate adoption of this technology in daily practice, we present an expert system based on the GPU-accelerated image processing library CLIJ: The CLIJ-assistant keeps track of which operations formed an image and suggests subsequent operations. It enables new ways of interaction with image data and image processing operations because its underlying GPU-accelerated image data flow graphs (IDFGs) allow changes to parameters of early processing steps and instantaneous visualization of their final results. Operations, their parameters and connections in the IDFG are stored at any point in time enabling the CLIJ-assistant to offer an undo-function for virtually unlimited rewinding parameter changes. Furthermore, to improve reproducibility of image data analysis workflows and interoperability with established image analysis platforms, the CLIJ-assistant can generate code from IDFGs in programming languages such as ImageJ Macro, Java, Jython, JavaScipt, Groovy, Python and C++ for later use in ImageJ, Fiji, Icy, Matlab, QuPath, Jupyter Notebooks and Napari. We demonstrate the CLIJ-assistant for processing image data in multiple scenarios to highlight its general applicability. The CLIJ-assistant is open source and available online: https://clij.github.io/assistant/

show abstract

“…We used a multi-resolution registration (2 resolutions) approach with b-spline interpolation (1st order in each resolution, 3rd order in final deformation), advanced Mattes mutual information similarity metric (32 histogram bins), advanced stochastic gradient descent optimizer (4000 max iterations), and b-spline transform (10 mm minimum grid spacing) 24 , 25 . T 1 maps were registered to the reference DCE-MRI frame using SimpleITK software 26 – 28 with a rigid transform model to match image matrix, voxel dimensions, and orientation. DWI images were first processed using FSL’s TopUp tool to correct susceptibility induced distortions 29 , 30 and the Eddy correction tool 31 .…”

Section: Methodsmentioning

confidence: 99%

Assessment of metastatic lymph nodes in head and neck squamous cell carcinomas using simultaneous 18F-FDG-PET and MRI

Chen

Hagiwara

Givi

et al. 2020

Sci Rep

View full text Add to dashboard Cite

In this study, we investigate the feasibility of using dynamic contrast enhanced magnetic resonance imaging (DCE-MRI), diffusion weighted imaging (DWI), and dynamic positron emission tomography (PET) for detection of metastatic lymph nodes in head and neck squamous cell carcinoma (HNSCC) cases. Twenty HNSCC patients scheduled for lymph node dissection underwent DCE-MRI, dynamic PET, and DWI using a PET-MR scanner within one week prior to their planned surgery. During surgery, resected nodes were labeled to identify their nodal levels and sent for routine clinical pathology evaluation. Quantitative parameters of metastatic and normal nodes were calculated from DCE-MRI (ve, vp, PS, Fp, Ktrans), DWI (ADC) and PET (Ki, K1, k2, k3) to assess if an individual or a combination of parameters can classify normal and metastatic lymph nodes accurately. There were 38 normal and 11 metastatic nodes covered by all three imaging methods and confirmed by pathology. 34% of all normal nodes had volumes greater than or equal to the smallest metastatic node while 4 normal nodes had SUV > 4.5. Among the MRI parameters, the median vp, Fp, PS, and Ktrans values of the metastatic lymph nodes were significantly lower (p = <0.05) than those of normal nodes. ve and ADC did not show any statistical significance. For the dynamic PET parameters, the metastatic nodes had significantly higher k3 (p value = 8.8 × 10−8) and Ki (p value = 5.3 × 10−8) than normal nodes. K1 and k2 did not show any statistically significant difference. Ki had the best separation with accuracy = 0.96 (sensitivity = 1, specificity = 0.95) using a cutoff of Ki = 5.3 × 10−3 mL/cm3/min, while k3 and volume had accuracy of 0.94 (sensitivity = 0.82, specificity = 0.97) and 0.90 (sensitivity = 0.64, specificity = 0.97) respectively. 100% accuracy can be achieved using a multivariate logistic regression model of MRI parameters after thresholding the data with Ki < 5.3 × 10−3 mL/cm3/min. The results of this preliminary study suggest that quantitative MRI may provide additional value in distinguishing metastatic nodes, particularly among small nodes, when used together with FDG-PET.

show abstract

Image Segmentation, Registration and Characterization in R with SimpleITK

Cited by 152 publications

References 40 publications

Split-Attention U-Net: A Fully Convolutional Network for Robust Multi-Label Segmentation from Brain MRI

Split-Attention U-Net: A Fully Convolutional Network for Robust Multi-Label Segmentation from Brain MRI

Interactive design of GPU-accelerated Image Data Flow Graphs and cross-platform deployment using multi-lingual code generation

Assessment of metastatic lymph nodes in head and neck squamous cell carcinomas using simultaneous 18F-FDG-PET and MRI

Contact Info

Product

Resources

About