NeRF를 이용한 실시간 볼륨 변형체 가시화
Realtime Volume Deformation Rendering using NeRF
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
본 연구는 의료 시술 후의 경과를 빠르고 간단하게 예측할 수 있는 가상 의료 계획 방법을 제안한다. 제안 방법은 촬영된 동영상으로부터 NeRF를 이용하여 볼륨데이터를 생성하고, 공간 분리를 통해 좌표계를 실공간과 가상공간으로 분리한다. 변형체 가시화는 가상공간에서의 역변환을 이용하여 수행한다. 또한, 사용자의 입력을 통해 변형하고자 하는 영역의 크기와 변형의 세기를 설정하여 다양한 교정 부위의 크기에 대해서 자연스러운 변형이 이루어지도록 한다. 변형은 변형 영역의 중심으로부터 멀어질수록 변형 세기가 감쇠되도록 좌표들 사이의 맨해튼 거리(manhattan distance)를 변형 세기의 가중치로 사용한다. 마지막으로, 그래픽스 처리장치(GPU)를 이용한 병렬화를 통해 대화적 시간으로 변형이 가능하다.
Abstract
This research proposes a virtual medical planning method that can quickly and simply predict the outcomes following medical procedures. The method generates volume data from recorded videos using Neural Radiance Fields (NeRF) and separates the coordinate system into real and virtual spaces through spatial separation. Visualization of deformations is performed using inverse transformations in the virtual space. Moreover, users can set the size and intensity of the desired deformations based on their input parameters, ensuring proper deformations for various sizes of correction areas. The intensity of the deformation diminishes as it moves away from the center of the deformation area, using the Manhattan distance between coordinates as a weight for the deformation intensity. Lastly, transformations can be performed in interactive time through parallel processing using a GPU.
참고문헌 (23)
- I. Badash, "Innovations in surgery simulation : a review of past, current and future techniques", Annals of Translational Medicine, 4(23), pp.453~ (2016)
- S. Gibson, "Simulating arthroscopic knee surgery using volumetric object representations, real-time volume rendering and haptic feedback", Medical Image Analysis, 2(2), pp.121~132 (1998)
- R. Bade, "Comparison of Fundamental Mesh Smoothing Algorithms for Medical Surface Models", Proceedings of Simulation and Visualization, 1, pp.289~304 (2006)
- M. Berger, "State of the art in surface reconstruction from point clouds", Proceedings of the 35th Annual Conference of the European Association for Computer Graphics, Eurographics 2014-State of the Art Reports, pp.161~185 (2014)
- A. Al-Khalifah, "Survey of Modeling Approaches for Medical Simulators", International Journal on Disability and Human Development, 4(3), pp.153~160 (2004)
- B. Mildenhall, "NeRF : Representing Scenes as Neural Radiance Fields for View Synthesis", Communications of the ACM, 65(1), pp.99~106 (2022)
- A. Nogales, "A Survey of Deep Learning Models in Medical Therapeutic Areas", Artificial Intelligence in Medicine, 112, pp.102020~ (2021)
- A. R. Aguilera, "A Parallel Resampling Method for Interactive Deformation of Volumetric Models", Computers & Graphics, 53, pp.147~155 (2015)
- J. Gascon, "Fast Deformation of Volume Data Using Tetrahedral Mesh Rasterization", Proceedings-SCA 2013: 12th ACM SIGGRAPH, 2013, 181~185
- F. Schulze, "Interactive Deformation and Visualization of Large Volume Datasets", GRAPP 2007 - 2nd International Conference on Computer Graphics Theory and Applications, Proceedings, pp.39~46 (2007)
- S. Lee, "A Study on the GPU Optimization of Deformation Using 3D Chainmail", Journal of Korea Multimedia Society, 26(2), pp.131~139 (2023)
- J. L. Schönberger, "Structurefrom-Motion Revisited", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.4104~4113 (2016)
- D. G. Lowe, "Distinctive Image Features from Scale-Invariant Keypoints", International Journal of Computer Vision, 60, pp.91~110 (2004)
- R. Hartley, "Multiple View Geometry in Computer Vision" (2003), Cambridge university press
- C. Beder, "Determining an Initial Image Pair for Fixing the Scale of a 3d Reconstruction from an Image Sequence", Joint Pattern Recognition Symposium, pp.657~666 (2006)
- M. A. Fischler, "Random Sample Consensus : A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography", Communications of the ACM, 24(6), pp.381~395 (1981)
- L. Kang, "Robust multi-view L2 triangulation via optimal inlier selection and 3d structure refinement", Pattern Recognition, 47(9), pp.2974~2992 (2014)
- B. Triggs, "Bundle Adjustment-A Modern Synthesis", Vision Algorithms : Theory and Practice, 1999 Proceedings, 1883, pp.298~372 (2000)
- T. Muller, "Instant neural graphics primitives with a multiresolution hash encoding", ACM Transactions on Graphics, 41(4), pp.1~15 (2022)
- J. Shin, "NeRFlex : Flexible Neural Radiance Fields with Diffeomorphic Deformation", IEEE Access, 12, pp.59920~59929 (2024)
- W. Deng, "Ray Deformation Networks for Novel View Synthesis of Refractive Objects", Proceedings of the IEEE, 2024, 3106~3116
- J. W. Seo, "GPU Parallelization Study of Position Based Dynamics for Medical Imaging", The Journal of KINGComputing, 19(3), pp.19~28 (2023)
- K. Y. Koo, "PBD Based Breast Region Deformation Simulation in AR Surgical Environments", The Journal of KINGComputing, 19(4), pp.76~84 (2023)