Hybrid 3D Mesh Reconstruction Models of CT Images for Deep Learning Based Classification of Kidney Tumors †
Engineering Proceedings, vol.104, no.1, 2025 (Scopus)
- Publication Type: Article / Article
- Volume: 104 Issue: 1
- Publication Date: 2025
- Doi Number: 10.3390/engproc2025104079
- Journal Name: Engineering Proceedings
- Journal Indexes: Scopus
- Keywords: anisotropic diffusion, kidney tumor segmentation, KiTS23 dataset, mesh reconstruction
- Kocaeli University Affiliated: Yes
Abstract
We present a comparative analysis of three hybrid methodologies for transforming 3D kidney tumor segmentations of volumetric NIfTI data into highly accurate network representations. Exploiting the KiTS23 dataset, we evaluate edge-preserving reconstruction pipelines integrating anisotropic diffusion, multiscale Gaussian filtering and KNN-based network optimisation. Model 1 uses Gaussian smoothing with Marching Cubes, while Model 2 uses spline interpolation and Perona-Malik filtering for improved resolution. Model 3 extends this structure with normal sensitive vertex smoothing to preserve critical anatomical interfaces. Quantitative metrics (Dice score, HD95) demonstrated the advantage of Model 3, which achieved a 22% reduction in the Hausdorff distance error rate compared to conventional methods while maintaining segmentation accuracy (Dice > 0.92). The proposed unsupervised pipeline bridges the gap between clinical interpretability and computational accuracy, providing a robust infrastructure for further applications in surgical planning and deep learning-based classification.