Parallel generation of digitally reconstructed radiographs on heterogeneous multi-GPU workstations

Abstract

The growing importance of three-dimensional radiotherapy treatment has been associated with the active presence of advanced computational workflows that can simulate conventional x-ray films from computed tomography (CT) volumetric data to create digitally reconstructed radiographs (DRR). These simulated x-ray images are used to continuously verify the patient alignment in image-guided therapies with 2D-3D image registration. The present DRR rendering pipelines are quite limited to handle huge imaging stacks generated by recent state-of-the-art CT imaging modalities. We present a high performance x-ray rendering pipeline that is capable of generating high quality DRRs from large scale CT volumes. The pipeline is designed to harness the immense computing power of all the heterogeneous computing platforms that are connected to the system relying on OpenCL. Load-balancing optimization is also addressed to equalize the rendering load across the entire system. The performance benchmarks demonstrate the capability of our pipeline to generate high quality DRRs from relatively large CT volumes at interactive frame rates using cost-effective multi-GPU workstations. A 5122 DRR frame can be rendered from 1024 × 2048 × 2048 CT volumes at 85 frames per second.

Intensity‐based 2D‐3D registration for an ACL reconstruction navigation system

“…As was shown in Figure , although splatting was a simple and fast rendering algorithm using only a small subset of voxels, the images generated by ray‐casting have been smoother than those generated by splatting. A multi‐GPU workstation was applied for effective generations of DRRs. Comparative results of different registration methods (Table ) showed that: splatting was faster than ray‐casting; the GD algorithm was more effective than DS, with less consumption time; and the registration accuracy of MMI was the worst, resulting in the only two failures of the registration.…”

X-ray system simulation software tools for radiology and radiography education

“…and are the input and output x-ray spectra respectively, as functions of photon energy expressed in electron volts, is the mass attenuation coefficient of the filtering material expressed in units cm 2 /g, is the density in units of g/cm 3 and is the thickness of the filtration in cm. Mass attenuation and density values were obtained from published tables [25].…”

“…An alternative to direct experimentation is the use of computer simulators. Existing techniques for the simulation of x-ray imaging systems have focussed primarily on Monte Carlo (MC) methods [1][2][3][4][5][6][7][8][9][10][11][12][13]. In general, these techniques are computationally expensive even when supported by graphical processing unit (GPU) acceleration [1][2][3][4][5][6][7] or when combined with deterministic calculations to speedup computation times [6,12,13].…”

“…Existing techniques for the simulation of x-ray imaging systems have focussed primarily on Monte Carlo (MC) methods [1][2][3][4][5][6][7][8][9][10][11][12][13]. In general, these techniques are computationally expensive even when supported by graphical processing unit (GPU) acceleration [1][2][3][4][5][6][7] or when combined with deterministic calculations to speedup computation times [6,12,13]. Other deterministic methods [14][15][16][17][18][19][20][21][22] employ ray or pencil beam tracing techniques using computed tomography (CT) volumes [14][15][16][17] or meshes [18,[20][21][22] to produce forward-projected radiographic 2D x-ray images.…”