.png){kind=icon}
返回
Validation Framework for Analyzing Complex Anatomical Structures: Application of L-System Models
.png){kind=icon}
OA
AI
DOI：10.1109/ACCESS.2025.3539837.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Accurate morphological and topological analysis of renal vascular trees is crucial for understanding vascular architecture, detecting patterns, and identifying pathological deviations. However, the complexity of renal vascular trees, particularly in corrosive endocasts, requires custom-designed algorithms to address unique challenges in reconstruction, segmentation, skeletonization, and graph-based representation. A key issue is the lack of a direct method to validate the correctness of these parameters in the absence of a ground truth prior to pattern analysis, with manual validation being infeasible due to the intricate nature of the renal vasculature. To address this challenge, we introduce a systematic validation framework utilizing artificially generated renal vascular models. We developed computational models using L-systems and fabricated them via 3D printing. These models were scanned using microcomputed tomography ( mu -CT), simulating the process for in vivo endocasts. We then applied the same computational pipeline to analyze the synthetic models, comparing the derived vascular parameters with the known geometries. Our results demonstrate that the computational methods can accurately replicate the morphology and topology of artificial models, with quantified validation showing high precision in the analysis. This approach provides a reliable methodology for validating vascular analysis algorithms, setting a benchmark for accurate morphological and topological assessments in complex vascular networks. It also highlights the value of synthetic models in verifying algorithm integrity, offering a solid foundation for future research on the analysis of complex anatomical structures.
Keyword:
Computational modeling
Solid modeling
Analytical models
Accuracy
Three-dimensional printing
Data models
Computed tomography
Image reconstruction
Imaging
Manuals
Validation framework
lack of ground truth
renal vascular trees
3D printing
microcomputed tomography (mu-CT)
L-systems
期刊
IF:
3.6
论文数:
9.8W
被引数:
29.4W
机构
引用论文
Unsupervised Segmentation of 3D Microvascular Photoacoustic Images Using Deep Generative Learning基于深度生成学习的三维微血管光声图像无监督分割.png)
ADVANCED SCIENCE
OAAI
ADVANCED SCIENCEIF14.1
OAAI分享收藏Symmetric diffeomorphic image registration with cross-correlation: Evaluating automated labeling of elderly and neurodegenerative brain具有互相关的对称微分图像配准: 评估老年人和神经退行性脑的自动标记MEDICAL IMAGE ANALYSISOAAI
MEDICAL IMAGE ANALYSISIF11.8
OAAI分享收藏Loading praziquantel within solid lipid nanoparticles improved its schistosomicidal efficacy against the juvenile stage将吡喹酮载入固体脂质纳米粒中提高了其对童虫期的杀血吸虫效力
PREAI
分享收藏VoxelMorph: A Learning Framework for Deformable Medical Image RegistrationVoxelMorph: 可变形医学图像配准的学习框架
PREAI
分享收藏Polar oxide substrates for graphene growth: A first-principles investigation of graphene on MgO(111)
PREAI
分享收藏
查看更多内容