CLAIRE: Constrained Large Deformation Diffeomorphic Image Registration on Parallel Computing Architectures

Brunn, Malte; Himthani, Naveen; Biros, George; Mehl, Miriam; Mang, Andreas

doi:10.21105/joss.03038

Cited by 5 publications

(11 citation statements)

References 23 publications

(40 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The scripts for generating the template image m 0 and the synthetic velocity field v can be found at https://github.com/naveenaero/scala-claire. The reference image m 1 can be generated using CLAIRE [15,66].…”

Section: Methodsmentioning

confidence: 99%

“…Our implementation uses PETSc [9,10] for linear algebra, the PETSc TAO package for nonlinear optimization, CUDA [77], thrust [44], cuFFT for FFTs [78], niftilib [31] for serial I/O for small images and PnetCDF [59] for parallel I/O for large scale images, IBM Spectrum MPI [2], and the IBM XL compiler [45]. [15,66] is available publicly for download on github at https://github.com/andreasmang/claire under the GNU General Public License v3.0.…”

Section: Compute Hardware and Librariesmentioning

confidence: 99%

“…There exist various algorithms and software packages for fast registration of images at standard clinical resolution (e.g., 256 3 ) [8,37,38,53,93,94,98,102]. This includes CLAIRE, which can execute image registration in parallel on multi-node multi-core CPUs and GPUs [14][15][16]35,67,69]. We note that there is little work on scalable image registration.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

CLAIRE -- Parallelized Diffeomorphic Image Registration for Large-Scale Biomedical Imaging Applications

Himthani¹,

Brunn²,

Kim³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

We study the performance of CLAIRE-a diffeomorphic multi-node, multi-GPU image registration algorithm and software-in large-scale biomedical imaging applications with billions of voxels. At such resolutions, most existing software packages for diffeomorphic image registration are prohibitively expensive. As a result, practitioners first significantly downsample the original images and then register them using existing tools. Our main contribution is an extensive analysis of the impact of downsampling on registration performance. We study this impact by comparing full resolution registrations obtained with CLAIRE to lower resolution registrations for synthetic and realworld imaging datasets. Our results suggest that registration at full resolution can yield a superior registration quality-but not always. For example, downsampling a synthetic image from 1024 3 to 256 3 decreases the Dice coefficient from 92% to 79%. However, the differences are less pronounced for noisy or low contrast high resolution images. CLAIRE allows us not only to register images of clinically relevant size in a few seconds but also to register images at unprecedented resolution in reasonable time. The highest resolution considered are CLARITY images of size 2816 × 3016 × 1162. To the best of our knowledge, this is the first study on image registration quality at such resolutions.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Compute Hardware and Librariesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

CLAIRE -- Parallelized Diffeomorphic Image Registration for Large-Scale Biomedical Imaging Applications

Himthani¹,

Brunn²,

Kim³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Different monitoring methods are selected for different soil types, and basic data monitoring is performed on the land in the uncontaminated area to obtain the soil environmental parameters in the original state of the area. For areas that have been polluted, soil sampling is performed directly to obtain initial land ecological pollution information data [18]. The monitoring equipment is arranged, and the soil profile is sampled in the monitoring area.…”

Section: Collection and Extraction Of Online Monitoringmentioning

confidence: 99%

Online Monitoring of Ecoenvironmental Pollution Data in Internet Environment

2022

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

In view of the problems of poor air quality and serious soil pollution in the current ecological environment, an online monitoring of ecological environment pollution data in the Internet environment is proposed. The concentration of pollution sources that may appear in the atmospheric environment is analyzed through the analysis coefficient of ecological environment pollutants, and the specific grade data corresponding to the air quality index is determined. Observation and mapping are carried out for the area to be monitored, and then, the monitoring points are determined, the monitoring equipment is arranged, and the soil profile of the area is sampled in the monitoring area. The self-organizing neural network model is used to conduct spatial cluster analysis on the pollution of ecological environment pollution areas, evaluate the sensitivity of soil pollution, and build a land ecological pollution monitoring information model. Based on the Internet of Things technology and the comprehensive use of GPS, sensors, and other technologies and equipment, the framework of monitoring technology is constructed, and the online monitoring of ecological environment pollution data is completed. The experimental results show that the proposed method has high data accuracy in different locations and different seasons, which is consistent with the actual situation; because of the maturity of the Internet of Things technology, compared with the literature method, the method in this paper can provide a cost-effective data foundation, which is more cost-effective, and save time and energy.

show abstract

“…There exist various algorithms and software packages for fast registration of images at standard clinical resolution (e.g., 256 3 ) [9][10][11][12][13][14][15][16]. This includes CLAIRE, which can execute image registration in parallel on multi-node multi-core CPUs and GPUs [17][18][19][20][21][22]. We note that there is little work on scalable image registration.…”

Section: Introductionmentioning

confidence: 99%

CLAIRE—Parallelized Diffeomorphic Image Registration for Large-Scale Biomedical Imaging Applications

et al. 2022

Self Cite

View full text Add to dashboard Cite

We study the performance of CLAIRE—a diffeomorphic multi-node, multi-GPU image-registration algorithm and software—in large-scale biomedical imaging applications with billions of voxels. At such resolutions, most existing software packages for diffeomorphic image registration are prohibitively expensive. As a result, practitioners first significantly downsample the original images and then register them using existing tools. Our main contribution is an extensive analysis of the impact of downsampling on registration performance. We study this impact by comparing full-resolution registrations obtained with CLAIRE to lower resolution registrations for synthetic and real-world imaging datasets. Our results suggest that registration at full resolution can yield a superior registration quality—but not always. For example, downsampling a synthetic image from 10243 to 2563 decreases the Dice coefficient from 92% to 79%. However, the differences are less pronounced for noisy or low contrast high resolution images. CLAIRE allows us not only to register images of clinically relevant size in a few seconds but also to register images at unprecedented resolution in reasonable time. The highest resolution considered are CLARITY images of size 2816×3016×1162. To the best of our knowledge, this is the first study on image registration quality at such resolutions.

show abstract

CLAIRE: Constrained Large Deformation Diffeomorphic Image Registration on Parallel Computing Architectures

Cited by 5 publications

References 23 publications

CLAIRE -- Parallelized Diffeomorphic Image Registration for Large-Scale Biomedical Imaging Applications

CLAIRE -- Parallelized Diffeomorphic Image Registration for Large-Scale Biomedical Imaging Applications

Online Monitoring of Ecoenvironmental Pollution Data in Internet Environment

CLAIRE—Parallelized Diffeomorphic Image Registration for Large-Scale Biomedical Imaging Applications

Contact Info

Product

Resources

About