PHASE IV AI HE project
Published June 30, 2025 | Version v1
Poster Open

Enhancing Medical Image Analysis: A Pipeline Combining Synthetic Image Generation and Super-Resolution

Authors/Creators

  • 1. ROR icon INESC TEC
  • 2. EDMO icon Faculty of Engineering - University of Porto (FEUP)
  • 3. Faculdade de Engenharia, Universidade do Porto
  • 4. ROR icon University of Coimbra
  • 5. Universidade do Porto Faculdade de Ciências

Description

Cancer is a leading cause of mortality worldwide, with breast and lung cancer being the most prevalent globally. Early and accurate diagnosis is crucial for successful treatment, and medical imaging techniques play a pivotal role in achieving this. This paper proposes a novel pipeline that leverages generative artificial intelligence to enhance medical images by combining synthetic image generation and super-resolution techniques. The framework is validated in two medical use cases (breast and lung cancers), demonstrating its potential to improve the quality and quantity of medical imaging data, ultimately contributing to more precise and effective cancer diagnosis and treatment. Overall, although some limitations do exist, this paper achieved satisfactory results for an image size which is conductive to specialist analysis, and further expands upon this field’s capabilities.

Files

[IbPRIA 2025] (Poster) Enhancing Medical Image Analysis A Pipeline Combining Synthetic Image Generation and Super-Resolution (1).pdf

Files (957.5 kB)

Additional details

Related works

Is supplement to
Conference paper: 10.5281/zenodo.15426164 (DOI)

References

  • I. Gulrajani, F. Ahmed, M. Arjovsky, V. Dumoulin, and A. Courville, 'Improved training of wasserstein GANs', Dec. 25, 2017, arXiv: arXiv:1704.00028.
  • G. Kwon, C. Han, and D. Kim, 'Generation of 3D brain MRI using auto-encoding generative adversarial networks', Aug. 07, 2019, arXiv: arXiv:1908.02498.
  • L. Sun, J. Chen, Y. Xu, M. Gong, K. Yu, and K. Batmanghelich, 'Hierarchical amortized GAN for 3D high resolution medical image synthesis', IEEE J. Biomed. Health Inform., vol. 26, no. 8, pp. 3966–3975, Aug. 2022, doi: 10.1109/JBHI.2022.3172976.
  • X. Wang, L. Xie, C. Dong, and Y. Shan, 'Real-ESRGAN: training real-world blind super-resolution with pure synthetic data', Aug. 17, 2021, arXiv: arXiv:2107.10833.