Memory-Efficient Super-Resolution of 3D Micro-CT Images Using Octree-Based GANs: Enhancing Resolution and Segmentation Accuracy

• URL: http://arxiv.org/abs/2505.18664v1
• Date: Sat, 24 May 2025 11:57:08 GMT
• Title: Memory-Efficient Super-Resolution of 3D Micro-CT Images Using Octree-Based GANs: Enhancing Resolution and Segmentation Accuracy
• Authors: Evgeny Ugolkov, Xupeng He, Hyung Kwak, Hussein Hoteit,
• Abstract summary: We present a memory-efficient algorithm for enhancing the quality of segmented 3D micro-d Tomography (micro-CT) images of rocks using a generative model.<n>The proposed model achieves a 16x increase in resolution and corrects inaccuracies in segmentation caused by the overlapping X-ray attenuation in micro-CT measurements across different minerals.<n>This framework demonstrates substantial improvements in pore characterization and mineral differentiation, offering a robust solution to one of the primary computational limitations in modern geoscientific imaging.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We present a memory-efficient algorithm for significantly enhancing the quality of segmented 3D micro-Computed Tomography (micro-CT) images of rocks using a generative model. The proposed model achieves a 16x increase in resolution and corrects inaccuracies in segmentation caused by the overlapping X-ray attenuation in micro-CT measurements across different minerals. The generative model employed is a 3D Octree-based convolutional Wasserstein generative adversarial network with gradient penalty. To address the challenge of high memory consumption inherent in standard 3D convolutional layers, we implemented an Octree structure within the 3D progressive growing generator model. This enabled the use of memory-efficient 3D Octree-based convolutional layers. The approach is pivotal in overcoming the long-standing memory bottleneck in volumetric deep learning, making it possible to reach 16x super-resolution in 3D, a scale that is challenging to attain due to cubic memory scaling. For training, we utilized segmented 3D low-resolution micro-CT images along with unpaired segmented complementary 2D high-resolution laser scanning microscope images. Post-training, resolution improved from 7 to 0.44 micro-m/voxel with accurate segmentation of constituent minerals. Validated on Berea sandstone, this framework demonstrates substantial improvements in pore characterization and mineral differentiation, offering a robust solution to one of the primary computational limitations in modern geoscientific imaging.

Related papers

• SparseFlex: High-Resolution and Arbitrary-Topology 3D Shape Modeling [79.56581753856452]
  SparseFlex is a novel sparse-structured isosurface representation that enables differentiable mesh reconstruction at resolutions up to $10243$ directly from rendering losses.<n>By enabling high-resolution, differentiable mesh reconstruction and generation with rendering losses, SparseFlex significantly advances the state-of-the-art in 3D shape representation and modeling.
  arXiv  Detail & Related papers  (2025-03-27T17:46:42Z)
• Self-Supervised Z-Slice Augmentation for 3D Bio-Imaging via Knowledge Distillation [65.46249968484794]
  ZAugNet is a fast, accurate, and self-supervised deep learning method for enhancing z-resolution in biological images.<n>By performing nonlinear distances between consecutive slices, ZAugNet effectively doubles resolution with each iteration.<n>ZAugNet+ is an extended version enabling continuous prediction at arbitrary distances.
  arXiv  Detail & Related papers  (2025-03-05T17:50:35Z)
• Super-Resolution of 3D Micro-CT Images Using Generative Adversarial Networks: Enhancing Resolution and Segmentation Accuracy [0.0]
  We develop a procedure for improving the quality of segmented 3D micro-CT images of rocks with a Machine Learning (ML) Generative Model.<n>The proposed model enhances the resolution eightfold (8x) and addresses segmentation inaccuracies due to the overlapping X-ray attenuation in micro-CT measurement for different rock minerals and phases.<n>We achieved high-quality super-resolved 3D images with a resolution of 0.4375 micro-m/voxel and accurate segmentation for constituting minerals and pore space.
  arXiv  Detail & Related papers  (2025-01-12T21:33:06Z)
• From Diffusion to Resolution: Leveraging 2D Diffusion Models for 3D Super-Resolution Task [19.56372155146739]
  We present a novel approach that leverages the 2D diffusion model and lateral continuity within the volume to enhance 3D volume electron microscopy (vEM) super-resolution. Our results on two publicly available focused ion beam scanning electron microscopy (FIB-SEM) datasets demonstrate the robustness and practical applicability of our framework.
  arXiv  Detail & Related papers  (2024-11-25T09:12:55Z)
• R$^2$-Gaussian: Rectifying Radiative Gaussian Splatting for Tomographic Reconstruction [53.19869886963333]
  3D Gaussian splatting (3DGS) has shown promising results in rendering image and surface reconstruction. This paper introduces R2$-Gaussian, the first 3DGS-based framework for sparse-view tomographic reconstruction.
  arXiv  Detail & Related papers  (2024-05-31T08:39:02Z)
• NeRF-GAN Distillation for Efficient 3D-Aware Generation with Convolutions [97.27105725738016]
  integration of Neural Radiance Fields (NeRFs) and generative models, such as Generative Adversarial Networks (GANs) has transformed 3D-aware generation from single-view images. We propose a simple and effective method, based on re-using the well-disentangled latent space of a pre-trained NeRF-GAN in a pose-conditioned convolutional network to directly generate 3D-consistent images corresponding to the underlying 3D representations.
  arXiv  Detail & Related papers  (2023-03-22T18:59:48Z)
• Towards Model Generalization for Monocular 3D Object Detection [57.25828870799331]
  We present an effective unified camera-generalized paradigm (CGP) for Mono3D object detection. We also propose the 2D-3D geometry-consistent object scaling strategy (GCOS) to bridge the gap via an instance-level augment. Our method called DGMono3D achieves remarkable performance on all evaluated datasets and surpasses the SoTA unsupervised domain adaptation scheme.
  arXiv  Detail & Related papers  (2022-05-23T23:05:07Z)
• Low dosage 3D volume fluorescence microscopy imaging using compressive sensing [0.0]
  We present a compressive sensing (CS) based approach to fully reconstruct 3D volumes with the same signal-to-noise ratio (SNR) with less than half of the excitation dosage. We demonstrate our technique by capturing a 3D volume of the RFP labeled neurons in the zebrafish embryo spinal cord with the axial sampling of 0.1um using a confocal microscope. The developed CS-based methodology in this work can be easily applied to other deep imaging modalities such as two-photon and light-sheet microscopy, where reducing sample photo-toxicity is a critical challenge.
  arXiv  Detail & Related papers  (2022-01-03T18:44:50Z)
• Deep learning of multi-resolution X-Ray micro-CT images for multi-scale modelling [0.0]
  We develop a 3D Enhanced Deep Super Resolution (EDSR) convolutional neural network to create enhanced, high-resolution data over large spatial scales. We validate the network with various metrics: textual analysis, segmentation behaviour and pore-network model (PNM) multiphase flow simulations. The EDSR generated model is more accurate than the base LR model at predicting experimental behaviour in the presence of heterogeneities.
  arXiv  Detail & Related papers  (2021-11-01T21:49:22Z)
• Revisiting 3D Context Modeling with Supervised Pre-training for Universal Lesion Detection in CT Slices [48.85784310158493]
  We propose a Modified Pseudo-3D Feature Pyramid Network (MP3D FPN) to efficiently extract 3D context enhanced 2D features for universal lesion detection in CT slices. With the novel pre-training method, the proposed MP3D FPN achieves state-of-the-art detection performance on the DeepLesion dataset. The proposed 3D pre-trained weights can potentially be used to boost the performance of other 3D medical image analysis tasks.
  arXiv  Detail & Related papers  (2020-12-16T07:11:16Z)
• Convolutional 3D to 2D Patch Conversion for Pixel-wise Glioma Segmentation in MRI Scans [22.60715394470069]
  We devise a novel pixel-wise segmentation framework through a convolutional 3D to 2D MR patch conversion model. In our architecture, both local inter-slice and global intra-slice features are jointly exploited to predict class label of the central voxel in a given patch.
  arXiv  Detail & Related papers  (2020-10-20T20:42:52Z)
• Hierarchical Amortized Training for Memory-efficient High Resolution 3D GAN [52.851990439671475]
  We propose a novel end-to-end GAN architecture that can generate high-resolution 3D images. We achieve this goal by using different configurations between training and inference. Experiments on 3D thorax CT and brain MRI demonstrate that our approach outperforms state of the art in image generation.
  arXiv  Detail & Related papers  (2020-08-05T02:33:04Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.