A Robust Fuzzy Ensemble Model For Cervical Cytology Image Classification

Early diagnosis and accurate classification of cervical cytology images are essential for making effective medical decisions. Cervix_dyk, Cervix_koc, Cervix_mep, Cervix_pab, and Cervix_sfi are the five categories into which cervical cytology images are divided in order to improve classification accuracy. In this work, a robust fuzzy ensemble model combining Inception_V3 and MobileNet_V2 is proposed. To successfully handle issues like uneven image quality and class imbalance, the suggested model makes use of the advantages of both deep learning architectures. According to experimental results, the fuzzy ensemble model achieves a 98% accuracy rate, outperforming standalone CNN models. This method facilitates more precise classification and diagnosis, which enhances clinical decision-making. The results highlight the promise of deep learning-based ensemble approaches for classifying medical images, providing insightful information for better diagnostic accuracy and early detection.