Project/Area Number |
16300006
|
Research Category |
Grant-in-Aid for Scientific Research (B)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Research Field |
Software
|
Research Institution | Osaka University |
Principal Investigator |
HAGIHARA Kenichi Osaka University, Graduate School of Information and Technology, Professor, 大学院・情報科学研究科, 教授 (00133140)
|
Co-Investigator(Kenkyū-buntansha) |
INO Fumihiko Osaka University, Graduate School of Information and Technology, Assistant Professor, 大学院・情報科学研究科, 助手 (90346172)
MIZUTANI Yasuharu Osaka Institute of Technology, Faculty of Information and Technology, Lecturer, 情報科学部, 講師 (10411414)
|
Project Period (FY) |
2004 – 2005
|
Project Status |
Completed (Fiscal Year 2005)
|
Budget Amount *help |
¥14,700,000 (Direct Cost: ¥14,700,000)
Fiscal Year 2005: ¥6,900,000 (Direct Cost: ¥6,900,000)
Fiscal Year 2004: ¥7,800,000 (Direct Cost: ¥7,800,000)
|
Keywords | computational GRID / medical image processing / high-performance computing / remote access / middleware for GRID / GPU / registration / GPUプログラミング / PCクラスタ / GPU装備PCクラスタ / データ圧縮 / グリッド / 非描画処理 |
Research Abstract |
This study dealt with the GRID middleware and parallel processing to realize image processing for medicine on computational GRID. 1. The following systems are developed. (1) A resource monitoring and selection method for rapid turnaround grid applications (for example, within 10 seconds) (2) A robust instrumentation method that is capable of generating large traces beyond the capacity of physical memory, aiming at analyzing the performance of large-scale parallel programs (3)A resource selection method for the GPU Grid, which utilizes the GPU as computational resources of the Grid (4) A robust instrumentation method that is capable of generating large traces beyond the capacity of physical memory, aiming at analyzing the performance of large-scale parallel programs 2.The following parallel algorithms are designed and implemented. (1) A nonrigid registration algorithm for investigating lung disease on PC clusters (2) A range of motion estimation method that is capable of fine-grained estimation during total hip replacement (THR) surgery (3) A biplane 2-D/3-D rigid registration using commodity Graphics Hardware (GPU) (4) A two-stage compression method for accelerating programmable GPU-based volume rendering of time-varying data (5) An efficient parallel volume rendering algorithm for visualization of large-scale datasets (6) An instruction allocation method that is capable of reducing the FP workload by moving some instructions from the FP program to the VP program, both written in a GPU assembly language. (7) An LU decomposition algorithm on the programmable GPU
|