Mul-material decomposition method for sandstone spectral CT images based on I-MultiEncFusion-Net
Material analysis in sandstone is essential for oil and gas extraction. Energy spectrum Computed Tomography (CT) can acquire various spectrally distinct datasets and reconstruct energy-selective images. Additionally, deep learning significantly improves the accuracy of material decomposition by esta...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-08-01
|
| Series: | Frontiers in Physics |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fphy.2025.1626220/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850028915149504512 |
|---|---|
| author | Yanfang Wu Yanfang Wu Yanfang Wu Ran Zhang Ran Zhang Ran Zhang Huihua Kong Huihua Kong Huihua Kong Ping Chen Ping Chen Ping Chen Yu Zou |
| author_facet | Yanfang Wu Yanfang Wu Yanfang Wu Ran Zhang Ran Zhang Ran Zhang Huihua Kong Huihua Kong Huihua Kong Ping Chen Ping Chen Ping Chen Yu Zou |
| author_sort | Yanfang Wu |
| collection | DOAJ |
| description | Material analysis in sandstone is essential for oil and gas extraction. Energy spectrum Computed Tomography (CT) can acquire various spectrally distinct datasets and reconstruct energy-selective images. Additionally, deep learning significantly improves the accuracy of material decomposition by establishing a nonlinear mapping relationship between multi-energy channel reconstructed images and their corresponding multi-material reconstructed images. However, traditional convolutional neural networks (CNNs) demonstrate limited effectiveness in capturing non-local features. In this paper, we present a multi-encoder single-decoder network architecture named I-MultiEncFusion-Net, designed for material decomposition. In this framework, multiple encoders concentrate on the distinctive features of reconstructed images from different energy spectrum channels, while a single decoder enables feature fusion. The encoder incorporates Inception_B modules that utilize three parallel branches to comprehensively capture image features, while integrating a Local-Nonlocal Feature Aggregation (LNFA) module to fuse both local and global characteristics. The non-local feature extraction module constructs non-local neighborhood relationships and employs Euclidean distance metrics to extract global contextual features from images, thereby enhancing the material decomposition process. To further enhance model accuracy, the decoder computes Huber loss between each output and its corresponding label, while simultaneously incorporating correlations of base material images extracted by a High-Resolution Network (HRNet) as an auxiliary loss constraint for material decomposition. Validation experiments using spectral CT data of sandstone demonstrate the method’s efficacy. Both simulated and practical results indicate that I-MultiEncFusion-Net exhibits superior generalization capability, preserves internal image details, and produces decomposed images with sharper edges. |
| format | Article |
| id | doaj-art-70a7108c24444226bcb0551bbab11edf |
| institution | DOAJ |
| issn | 2296-424X |
| language | English |
| publishDate | 2025-08-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Physics |
| spelling | doaj-art-70a7108c24444226bcb0551bbab11edf2025-08-20T02:59:41ZengFrontiers Media S.A.Frontiers in Physics2296-424X2025-08-011310.3389/fphy.2025.16262201626220Mul-material decomposition method for sandstone spectral CT images based on I-MultiEncFusion-NetYanfang Wu0Yanfang Wu1Yanfang Wu2Ran Zhang3Ran Zhang4Ran Zhang5Huihua Kong6Huihua Kong7Huihua Kong8Ping Chen9Ping Chen10Ping Chen11Yu Zou12School of Mathematics, North University of China, Taiyuan, ChinaNational Key Laboratory of Photoelectric Dynamic Testing Technology and Instrument for Extreme Environments, North University of China, Taiyuan, ChinaShanxi Key Laboratory of Signal Capturing and Processing, North University of China, Taiyuan, ChinaSchool of Mathematics, North University of China, Taiyuan, ChinaNational Key Laboratory of Photoelectric Dynamic Testing Technology and Instrument for Extreme Environments, North University of China, Taiyuan, ChinaShanxi Key Laboratory of Signal Capturing and Processing, North University of China, Taiyuan, ChinaSchool of Mathematics, North University of China, Taiyuan, ChinaNational Key Laboratory of Photoelectric Dynamic Testing Technology and Instrument for Extreme Environments, North University of China, Taiyuan, ChinaShanxi Key Laboratory of Signal Capturing and Processing, North University of China, Taiyuan, ChinaNational Key Laboratory of Photoelectric Dynamic Testing Technology and Instrument for Extreme Environments, North University of China, Taiyuan, ChinaShanxi Key Laboratory of Signal Capturing and Processing, North University of China, Taiyuan, ChinaSchool of Information and Communication Engineering, Taiyuan, ChinaState Key Laboratory of Lithospheric and Environmental Coevolution, Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing, ChinaMaterial analysis in sandstone is essential for oil and gas extraction. Energy spectrum Computed Tomography (CT) can acquire various spectrally distinct datasets and reconstruct energy-selective images. Additionally, deep learning significantly improves the accuracy of material decomposition by establishing a nonlinear mapping relationship between multi-energy channel reconstructed images and their corresponding multi-material reconstructed images. However, traditional convolutional neural networks (CNNs) demonstrate limited effectiveness in capturing non-local features. In this paper, we present a multi-encoder single-decoder network architecture named I-MultiEncFusion-Net, designed for material decomposition. In this framework, multiple encoders concentrate on the distinctive features of reconstructed images from different energy spectrum channels, while a single decoder enables feature fusion. The encoder incorporates Inception_B modules that utilize three parallel branches to comprehensively capture image features, while integrating a Local-Nonlocal Feature Aggregation (LNFA) module to fuse both local and global characteristics. The non-local feature extraction module constructs non-local neighborhood relationships and employs Euclidean distance metrics to extract global contextual features from images, thereby enhancing the material decomposition process. To further enhance model accuracy, the decoder computes Huber loss between each output and its corresponding label, while simultaneously incorporating correlations of base material images extracted by a High-Resolution Network (HRNet) as an auxiliary loss constraint for material decomposition. Validation experiments using spectral CT data of sandstone demonstrate the method’s efficacy. Both simulated and practical results indicate that I-MultiEncFusion-Net exhibits superior generalization capability, preserves internal image details, and produces decomposed images with sharper edges.https://www.frontiersin.org/articles/10.3389/fphy.2025.1626220/fulli-MultiEncFusion-nethigh-resolution networkmulti-material decompositionlayer normalization and feature aggregationenergy spectrum CT |
| spellingShingle | Yanfang Wu Yanfang Wu Yanfang Wu Ran Zhang Ran Zhang Ran Zhang Huihua Kong Huihua Kong Huihua Kong Ping Chen Ping Chen Ping Chen Yu Zou Mul-material decomposition method for sandstone spectral CT images based on I-MultiEncFusion-Net Frontiers in Physics i-MultiEncFusion-net high-resolution network multi-material decomposition layer normalization and feature aggregation energy spectrum CT |
| title | Mul-material decomposition method for sandstone spectral CT images based on I-MultiEncFusion-Net |
| title_full | Mul-material decomposition method for sandstone spectral CT images based on I-MultiEncFusion-Net |
| title_fullStr | Mul-material decomposition method for sandstone spectral CT images based on I-MultiEncFusion-Net |
| title_full_unstemmed | Mul-material decomposition method for sandstone spectral CT images based on I-MultiEncFusion-Net |
| title_short | Mul-material decomposition method for sandstone spectral CT images based on I-MultiEncFusion-Net |
| title_sort | mul material decomposition method for sandstone spectral ct images based on i multiencfusion net |
| topic | i-MultiEncFusion-net high-resolution network multi-material decomposition layer normalization and feature aggregation energy spectrum CT |
| url | https://www.frontiersin.org/articles/10.3389/fphy.2025.1626220/full |
| work_keys_str_mv | AT yanfangwu mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT yanfangwu mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT yanfangwu mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT ranzhang mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT ranzhang mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT ranzhang mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT huihuakong mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT huihuakong mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT huihuakong mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT pingchen mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT pingchen mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT pingchen mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet AT yuzou mulmaterialdecompositionmethodforsandstonespectralctimagesbasedonimultiencfusionnet |