DEVELOPMENT OF AN ARTIFICIAL INTELLIGENCE-BASED PATTERN RECOGNITION MODEL FOR EARLY DISEASE DETECTION
Keywords:
artificial intelligence, pattern recognition, early disease detection, deep learning, convolutional neural networks, attention mechanism, medical image analysis, computer-aided diagnosis, feature extraction.artificial intelligence, pattern recognition, early disease detection, deep learning, convolutional neural networks, attention mechanism, medical image analysis, computer-aided diagnosis, feature extraction.Abstract
This article addresses the development and analysis of artificial intelligence (AI) and deep learning models designed for the early diagnosis of diseases based on medical imaging and biometric data. The primary objective of the research is to detect subtle pathological changes with high accuracy and a minimal error rate using convolutional neural networks (CNNs) and hybrid architectures. Methods for data preprocessing, noise reduction, feature extraction and model optimization are proposed. The obtained results demonstrate a significant increase in efficiency and speed compared to conventional diagnostic methods: the proposed hybrid CNN + Attention model reached an accuracy of 96.8% on the test set.
References
1. Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention (MICCAI 2015) (pp. 234–241). Springer.
2. Applied and Computational Engineering Editorial Board. (2025). Deep learning in medical image analysis. Applied and Computational Engineering, 210(1), 179–185. https://doi.org/10.54254/2755-2721/2025.LD30225
3. Palaniappan, D., et al. (2025). YOLO in healthcare: A comprehensive review of detection architectures, domain applications, and future innovations. IEEE Access, 13, 145714–145735. https://doi.org/10.1109/ACCESS.2025.3599358
4. Research Staff. (2025). Deep learning-based medical object detection: A survey. IEEE Access, 13, 53019–53038. https://doi.org/10.1109/ACCESS.2025.3553087
