Robust Joint Detection of Coronary Artery Plaque and Stenosis in Angiography Using Enhanced DCNN-GAN

Timely detection and diagnosis of coronary artery segment plaque and stenosis in X-ray angiography is of great significance, however, the image quality variation, noise, and artifacts in the original image cause definitive difficulties to the current algorithms. These problems pose a challenge to meaningful analysis via traditional approaches, which compromises the efficiency of detection algorithms. To overcome these drawbacks, the current study presents a new integrated deep learning technique that integrates Deep Convolutional Neural Network (DCNN) with Generative Adversarial Network (GAN) in dual conditional detection. Detailed feature learning extracted from X-ray angiography images are performed through DCNN where it considers vascular structure and automatic pathologic regions detection. The use of GANs is to further enrich the dataset with synthetic images, distortions, and visual noise, which will make the model more immune to various conditions of images. Both approaches combined help in better classification of normal and pathological areas and less sensitiveness to quality of the obtained images. The proposed method therefore has shown an improvement of the diagnostic accuracy as a solid foundation for clinical decision making in cardiovascular systems. The efficacy of the suggested approach has been demonstrated by the following evaluation metrics: 97.9% F1 score, 98.7% accuracy, 98.2% precision, and 98% recall. The results prove higher sensitivity and accuracy of the plaque and stenosis identification comparing to the traditional methods, which confirms the efficiency of using the proposed DCNN-GAN method for considering the real-world fluctuations in the medical imaging. It reveals a decisive advancement in the ability to use algorithms for cardiovascular assessment by providing better results in difficult imaging environments.

@article{Jayasree2025,
  title     = {Robust Joint Detection of Coronary Artery Plaque and Stenosis in Angiography Using Enhanced DCNN-GAN},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {16},
  number    = {1},
  year      = {2025},
  publisher = {The Science and Information Organization},
  author    = {M. Jayasree and L. Koteswara Rao},
  doi       = {10.14569/IJACSA.2025.0160195},
  url       = {https://doi.org/10.14569/IJACSA.2025.0160195}
}