Improving Quality of Medical Scans using GANs


Authors : Tanushree Bharti; Yogam Singh; Mudit Jain; Ankita Kumari

Volume/Issue : Volume 9 - 2024, Issue 12 - December

Google Scholar : https://tinyurl.com/ms68734e

Scribd : https://tinyurl.com/3wnsa3vt

DOI : https://doi.org/10.5281/zenodo.14557003

Abstract : Improving the quality of medical images is essential for precise diagnosis and treatment planning. When low quality images are used to train the neural network model, the good accuracy cannot be achieved. Nowadays, Generative Adversarial Networks (GANs) have become a potent image enhancement tool that can provide a fresh method for raising the caliber of medical images. In order to improve medical images, this paper presents a GAN-based framework that reduces noise, increases resolution, and corrects artifacts. The suggested technique makes use of a generator network to convert low-quality images into their high-quality equivalents, and a discriminator network to assess the veracity of the improved images. To ensure robustness across various modalities, the model is trained on a diverse dataset of medical images, including MRI, CT, and X-ray scans. Our experimental results show that GAN-based method significantly improves the image quality when compared to conventional methods, as evidenced by enhanced peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM) according to quantitative evaluations. This study emphasizes the value of incorporating deep learning methods into medical image processing pipelines and the potential of GANs to advance medical imaging technology so that a robust neural network model can be designed.

Keywords : Medical Images Quality, Convolutional Neural Networks, Generative Adversarial Networks (GANs), Peak Signal to Noise Ratio (PSNR), Structural Similarity Index (SSIM).

References :

  1. Hoo-Chang Shin et al. "Synthetic data augmentation using GAN for improved liver lesion classification." In: International Workshop on Medical Imaging Simulation and Synthesis, pages 1-11. Cham, Springer, 2018.
  2. Eunji Choi et al. "Stargan: Unified generative adversarial networks for multi-domain image-to-image translation." In 2018's IEEE Conference on Computer Vision and Pattern Recognition Proceedings, pp. 8789–8797.
  3. Jelmer M. Wolterink and colleagues "Generative adversarial networks for noise reduction in low-dose CT." IEEE Medical Imaging Transactions 36.12, 2017; 2536–2545.
  4. Maayan Frid-Adar et al. "Synthetic data augmentation using GAN for improved liver lesion classification." In: International Workshop on Medical Imaging Simulation and Synthesis, pages 1-11. Cham, Springer, 2018.
  5. Armanious, Kristina, et al. "Using adversarial networks to synthesize CT from ultrasound images." In International Conference on Computer-Assisted Intervention and Medical Image Computing, pp. 81–89. Springer, Cham (2017).
  6. Qiaoying Yang et al. "Low-dose CT image denoising using a generative adversarial network with Wasserstein distance and perceptual loss." 2018 IEEE Access 6: 47958–47966.
  7. Chunyuan Li and colleagues, "Unsupervised image-to- image translation networks." In Neural Information Processing Systems Advances, pp. 700–708 in 2017.
  8. Yibin Song and colleagues, "Liver lesion detection and classification with novel neural network architectures." In: International Conference on Computer-Assisted Intervention and Medical Image Computing, 830-838. Springer, Cham (2017).
  9. Zhu, Jun-Yan, and associates. "Toward multimodal image-to-image translation." Pages 465–476 in Advances in Neural Information Processing Systems, 2017.
  10. Zhang, et al., "Translating and segmenting multimodal medical volumes with cycle-and shape-consistency generative adversarial network." In International Conference on Computer-Assisted Intervention and Medical Image Computing, pages 56–64. Cham, Springer, 2018.
  11. Kavita Lal, Madan Lal Saini; A study on deep fake identification techniques using deep learning. AIP Conf. Proc. 15 June 2023; 2782 (1): 020155. https://doi.org/10.1063/5.0154828
  12. Y. Singh, M. Saini and Savita, "Impact and Performance Analysis of Various Activation Functions for Classification Problems," 2023 IEEE International Conference on Contemporary Computing and Communications (InC4), Bangalore, India, 2023, pp. 1-7, doi: 10.1109/InC457730.2023.10263129.
  13. Sarmah, J., Saini, M.L., Kumar, A., Chasta, V. (2024). Performance Analysis of Deep CNN, YOLO, and LeNet for Handwritten Digit Classification. In: Sharma, H., Chakravorty, A., Hussain, S., Kumari, R. (eds) Artificial Intelligence: Theory and Applications. AITA 2023. Lecture Notes in Networks and Systems, vol 844. Springer, Singapore. https://doi.org/10.1007/978-981-99-8479-4_16
  14. M. L. Saini, A. Patnaik, Mahadev, D. C. Sati and R. Kumar, "Deepfake Detection System Using Deep Neural Networks," 2024 2nd International Conference on Computer, Communication and Control (IC4), Indore, India, 2024, pp. 1-5, doi: 10.1109/IC457434.2024.10486659.
  15. P. D. S. Prasad, R. Tiwari, M. L. Saini and Savita, "Digital Image Enhancement using Conventional Neural Network," 2023 2nd International Conference for Innovation in Technology (INOCON), Bangalore, India, 2023, pp. 1-5, doi: 10.1109/INOCON57975.2023.10100995.
  16. Skandarani, Youssef, Pierre-Marc Jodoin, and Alain Lalande. "Gans for medical image synthesis: An empirical study." Journal of Imaging 9.3 (2023): 69.
  17. Showrov, Atif Ahmed, et al. "Generative Adversarial Networks (GANs) in Medical Imaging: Advancements, Applications and Challenges." IEEE Access (2024).
  18. E. G. Kumar, M. Lal Saini, S. A. Khadar Ali and B. B. Teja, "A Clinical Support System for Prediction of Heart Disease using Ensemble Learning Techniques," 2023 International Conference on Sustainable Communication Networks and Application (ICSCNA), Theni, India, 2023, pp. 926-931, doi: 10.1109/ICSCNA58489.2023.10370569

Improving the quality of medical images is essential for precise diagnosis and treatment planning. When low quality images are used to train the neural network model, the good accuracy cannot be achieved. Nowadays, Generative Adversarial Networks (GANs) have become a potent image enhancement tool that can provide a fresh method for raising the caliber of medical images. In order to improve medical images, this paper presents a GAN-based framework that reduces noise, increases resolution, and corrects artifacts. The suggested technique makes use of a generator network to convert low-quality images into their high-quality equivalents, and a discriminator network to assess the veracity of the improved images. To ensure robustness across various modalities, the model is trained on a diverse dataset of medical images, including MRI, CT, and X-ray scans. Our experimental results show that GAN-based method significantly improves the image quality when compared to conventional methods, as evidenced by enhanced peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM) according to quantitative evaluations. This study emphasizes the value of incorporating deep learning methods into medical image processing pipelines and the potential of GANs to advance medical imaging technology so that a robust neural network model can be designed.

Keywords : Medical Images Quality, Convolutional Neural Networks, Generative Adversarial Networks (GANs), Peak Signal to Noise Ratio (PSNR), Structural Similarity Index (SSIM).

Never miss an update from Papermashup

Get notified about the latest tutorials and downloads.

Subscribe by Email

Get alerts directly into your inbox after each post and stay updated.
Subscribe
OR

Subscribe by RSS

Add our RSS to your feedreader to get regular updates from us.
Subscribe