Autors: Stefanov H., Zapryanov, G. S., Nikolova, I. N.
Title: Genetically Optimized Architectural Design of CNN for Medical CT Image Processing
Keywords: Convolutional Neural Networks (CNN), CT Image Processing, Genetic Algorithm, Medical Diagnostics, Neural Architecture Optimization

Abstract: Since the emergence of deep learning, optimizing neural network architectures has become a key factor in improving medical image analysis. This paper presents a genetic algorithm for optimizing CNN architectures in the classification of medical CT images, focusing on COVID-19related lung inflammation. The approach includes a dynamic layer addition mechanism that allows model complexity to adapt during training. Experiments on the COVID-CTx dataset demonstrate improved accuracy compared to random search, highlighting the potential of evolutionary methods for automated neural architecture design in medical imaging tasks.

References

  1. Litjens, Geert, et al. "A survey on deep learning in medical image analysis." Medical image analysis 42 (2017): 60-88.
  2. Bischl, Bernd, et al. "Hyperparameter optimization: Foundations, algorithms, best practices, and open challenges." Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery 13.2 (2023): e1484.
  3. Ali, Muhammad Junaid, et al. "A review of AutoML optimization techniques for medical image applications." Computerized Medical Imaging and Graphics (2024): 102441.
  4. Wolpert, David H., and William G. Macready. "No free lunch theorems for optimization." IEEE transactions on evolutionary computation 1.1 (1997): 67-82.
  5. Bergstra, James, and Yoshua Bengio. "Random search for hyper-parameter optimization." The journal of machine learning research 13.1 (2012): 281-305.
  6. Real, E., Aggarwal, A., Huang, Y., & Le, Q. V. (2019). Regularized Evolution for Image Classifier Architecture Search. Proceedings of the AAAI Conference on Artificial Intelligence, 33(01), 4780-4789.
  7. "COVID‑CTx dataset." Kaggle, n.d. [Online]. Available. https://www.kaggle.com/datasets/hgunraj/covidxct.
  8. Xiao, Xueli, et al. "Efficient hyperparameter optimization in deep learning using a variable length genetic algorithm." arXiv preprint arXiv:2006.12703 (2020).

Issue

2025 13th International Scientific Conference on Computer Science, COMSCI 2025 - Proceedings, 2025, Bulgaria, https://doi.org/10.1109/COMSCI67172.2025.11225238

Вид: публикация в международен форум, публикация в реферирано издание, индексирана в Scopus