Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{204211,
        author = {Sana Afrin and Purva Patil and Sejal Pawar and Tushar Mahajan},
        title = {Deep Learning-Based HEp-2 Cell Classification for Autoimmune Disease Detection Using Transfer Learning},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {3003-3009},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=204211},
        abstract = {Autoimmune diseases occur when the immune system wrongly attacks healthy tissues and organs. This causes chronic inflammation and serious health problems. Early diagnosis of autoimmune disorders is crucial for effective treatment and disease management. The Indirect Immunofluorescence (IIF) test on Human Epithelial Type-2 (HEp-2) cells is commonly used to detect Antinuclear Antibodies (ANA), which are key markers of autoimmune diseases. However, manually interpreting HEp-2 cell patterns takes a lot of time, can be subjective, and may vary from one observer to another. This research introduces an automated system for classifying HEp-2 cells using Transfer Learning and Deep Learning techniques. The proposed framework uses the MobileNetV2 architecture, which was pretrained on ImageNet, as a feature extractor for classifying six major HEp-2 staining patterns: Homogeneous, Speckled, Nucleolar, Centromere, Cytoplasmic, and Golgi. Image preprocessing techniques, such as denoising, contrast improvement using CLAHE, resizing, and normalization, are applied to enhance image quality and model performance. Data enhancement and fine-tuning are used to improve generalization and classification accuracy. Additionally, a confidence-based normal cell detection mechanism is introduced. Predictions that fall below a set confidence threshold are labeled as Normal or Unknown, which lowers the risk of false disease predictions. The experimental results show that the proposed approach effectively classifies HEp-2 cell patterns and aids in diagnosing autoimmune diseases.},
        keywords = {Autoimmune Disease Detection, HEp-2 Cells, Transfer Learning, MobileNetV2, Deep Learning, Medical Image Classification, Biomedical Imaging, Computer-Aided Diagnosis.},
        month = {June},
        }