Medical Image Fusion Using Symmetric and Asymmetric Strategies: Evaluating Classical and Deep Learning Methods
Received: 22 October 2025 | Revised: 14 November 2025 and 27 November 2025 | Accepted: 28 November 2025 | Online: 20 December 2025
Corresponding author: Hayath T M
Abstract
By combining complementary data from various imaging modalities, including Computed Tomography (CT) and Magnetic Resonance Imaging (MRI), multimodal medical image fusion significantly improves diagnostic accuracy. Despite notable progress, there are still no comprehensive assessment frameworks that fairly contrast cutting-edge deep learning methods with conventional symmetric and asymmetric fusion techniques. Using a wide range of models, such as Average Fusion, Principal Component Analysis (PCA), Discrete Wavelet Transform (DWT), Laplacian Pyramid, Saliency Mapping, Guided Filtering, and the DenseFuse Convolutional Neural Network (CNN) framework, this study suggests a unified comparative method for fusion of multimodal medical images. All techniques were applied to uniformly preprocessed CT-MRI datasets and assessed using a variety of quantitative metrics, including the Edge Preservation Index (EPI), Mean Squared Error (MSE), Entropy, Mutual Information (MI), Standard Deviation (STD), Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index (SSIM). The findings show that although deep learning-based methods such as DenseFuse provide strong feature extraction capabilities, they may have stability and generalizability issues in specific situations. However, asymmetric techniques provided better edge preservation and localized detail enhancement, while conventional symmetric techniques such as PCA and DWT demonstrated more consistent and interpretable performance across metrics. In addition to outlining the advantages and disadvantages of each category, this thorough analysis offers vital information to select the best fusion technique based on image properties, computational efficiency, and diagnostic requirements. The findings show that PCA and Guided Filtering produced superior SSIM and PSNR values (up to 0.9766 and 35.16 dB, respectively), while deep learning-based DenseFuse showed performance limitations under the given conditions, indicating domain adaptation issues. The results provide a strong basis for further investigation into multimodal image fusion and its useful implementation in clinical settings.
Keywords:
medical image fusion, multimodal imaging, CT-MRI fusion, symmetric fusion, asymmetric fusion, deep learning, DenseFuse, Principal Component Analysis (PCA), Discrete Wavelet Transform (DWT), Laplacian pyramid, guided filtering, saliency mapping, performance evaluation metrics, Structural Similarity Index (SSIM), Peak Signal-to-Noise Ratio (PSNR)Downloads
References
K. Vanitha, D. Satyanarayana, and M. N. GiriPrasad, "A novel framework of multimodal medical image fusion using adaptive NSST and optimized deep learning approach," The Imaging Science Journal, vol. 72, no. 8, pp. 1015–1042, Nov. 2024. DOI: https://doi.org/10.1080/13682199.2023.2241793
C. Ghandour, W. El-Shafai, and S. El-Rabaie, "Medical Image Fusion Based on Weighted Least Square Optimization and Deep Learning Algorithm," in 2021 9th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC), Alexandria, Egypt, Dec. 2021, pp. 159–163. DOI: https://doi.org/10.1109/JAC-ECC54461.2021.9691453
Z. Senousy, M. M. Gaber, and M. M. Abdelsamea, "AUQantO: Actionable Uncertainty Quantification Optimization in deep learning architectures for medical image classification," Applied Soft Computing, vol. 146, Oct. 2023, Art. no. 110666. DOI: https://doi.org/10.1016/j.asoc.2023.110666
D. M. Migayo, S. Kaijage, S. Swetala, and D. G. Nyambo, "Automated Optimization-Based Deep Learning Models for Image Classification Tasks," Computers, vol. 12, no. 9, Sept. 2023, Art. no. 174. DOI: https://doi.org/10.3390/computers12090174
P. Tata and M. A. Sowjnaya, "Hybrid optimization enabled Random multimodal deep learning for sentiment rating prediction," Intelligent Decision Technologies, vol. 18, no. 2, pp. 965–979, June 2024. DOI: https://doi.org/10.3233/IDT-220036
Z. Al‐Milaji and H. Yousif, "Lightweight Deep Learning Model Optimization for Medical Image Analysis," International Journal of Imaging Systems and Technology, vol. 34, no. 5, Sept. 2024, Art. no. e23173. DOI: https://doi.org/10.1002/ima.23173
G. Ma and Y. Li, "Optimization Research of Deep Learning Algorithms in Real-time Image Processing," Journal of Higher Education Research, vol. 5, no. 5, Nov. 2024, Art. no. 476. DOI: https://doi.org/10.32629/jher.v5i5.3065
G. Ferreira, M. H. D. N. Marinho, V. Severo, and F. Madeiro, "Optimization Strategies Applied to Deep Learning Models for Image Steganalysis: Application of Pruning, Quantization and Weight Clustering," Applied Sciences, vol. 15, no. 9, Apr. 2025, Art. no. 4632. DOI: https://doi.org/10.3390/app15094632
D. Shulman, "Optimization Methods in Deep Learning: A Comprehensive Overview." arXiv, 2023.
S. I. Nikolenko, "Deep Learning and Optimization," in Synthetic Data for Deep Learning, vol. 174, Springer International Publishing, 2021, pp. 19–58. DOI: https://doi.org/10.1007/978-3-030-75178-4_2
M. Wei, M. Xi, Y. Li, M. Liang, and G. Wang, "Multimodal Medical Image Fusion: The Perspective of Deep Learning," Academic Journal of Science and Technology, vol. 5, no. 3, pp. 202–208, May 2023. DOI: https://doi.org/10.54097/ajst.v5i3.8013
A. Gopatoti, K. K. Gopathoti, and N. R. Gujjula, Deep Learning for Medical Image Segmentation and Analysis. GSE Publications, 2024. DOI: https://doi.org/10.58599/9788197330537.15052024
Y. Li et al., "A review of deep learning-based information fusion techniques for multimodal medical image classification." arXiv, 2024.
S. Kingsley and T. Sethukarasi, "Optimization enabled deep learning approach with probabilistic fusion for image inpainting," The Imaging Science Journal, vol. 72, no. 5, pp. 591–614, July 2024. DOI: https://doi.org/10.1080/13682199.2023.2216976
A. Vylala, B. P. Radhakrishnan, and A. B. Kadan, "Early Prediction and Risk Analysis Using Hybrid Deep Learning Techniques in Multimodal Biomedical Image," Developmental Neurobiology, vol. 85, no. 4, Oct. 2025, Art. no. e23001. DOI: https://doi.org/10.1002/dneu.23001
H. Yan, Z. Wang, Y. Zhao, Y. Zhang, and R. Lyu, "Research on Image Generation Optimization based Deep Learning." Computer Science and Mathematics, Dec. 30, 2024. DOI: https://doi.org/10.20944/preprints202408.0927.v2
H. Yan, Z. Wang, S. Bo, Y. Zhao, Y. Zhang, and R. Lyu, "Research on Image Generation Optimization based Deep Learning." Computer Science and Mathematics, Aug. 13, 2024. DOI: https://doi.org/10.20944/preprints202408.0927.v1
Y. N. G and Dr. S. M. Kumar, "A Simple Hybrid Image Segmentation Approach Combining Classical And Deep Learning Techniques For Medical And General Image Analysis," IJRDO - Journal of Computer Science Engineering, 2025.
H. Keawmuang, S. Hu, T. Badloe, S. So, and J. Rho, "Hybrid Frameworks Integrating Deep Learning and Optimization Methods for Inverse Design in Nanophotonics," ACS Applied Materials & Interfaces, vol. 17, no. 23, pp. 33259–33270, June 2025. DOI: https://doi.org/10.1021/acsami.5c03196
S. Luo, "A Survey on Multimodal Deep Learning for Image Synthesis: Applications, methods, datasets, evaluation metrics, and results comparison," in 2021 the 5th International Conference on Innovation in Artificial Intelligence, Xia men, China, Mar. 2021, pp. 108–120. DOI: https://doi.org/10.1145/3461353.3461388
M. A. Al-Ebrahim, "Spike-Based Attention Mechanisms for Enhanced Medical Image Segmentation," Engineering, Technology & Applied Science Research, vol. 15, no. 5, pp. 28273–28285, Oct. 2025. DOI: https://doi.org/10.48084/etasr.13407
"BRaTS 2021 Task 1 Dataset." [Online]. Available: https://www.kaggle.com/datasets/dschettler8845/brats-2021-task1.
Downloads
How to Cite
License
Copyright (c) 2025 T. M. Hayath, Sai D. Madhavi
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain the copyright and grant the journal the right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) after its publication in ETASR with an acknowledgement of its initial publication in this journal.
