Enhancing Cervical Cancer Diagnosis: Deep Learning and Explainable AI in Multiclass Image Classification

Cervical Cancer is a major health concern amongst many women around the world. Cervical Cancer is usually caused by the Human Papilloma Virus (HPV). This causes unusual growth in the cervix region. All though this is such a serious health concern, proper treatment and diagnosis are not that prominent and available to the patients. Regular checkups and tests are quite necessary which can help in early detection and diagnosis. It has been observed if these cancerous growths are detected at an early stage then the patient has a high recovery rate. If not detected early then there are chances that these cancer cells spread to the other organs which can be a fatal condition for the patient. A technique has been introduced involving the use of deep learning models to categorize cancer cells into four distinct classes, aiding in accurate identification and effective diagnosis. Explainable AI techniques like Local Interpretable Model Agnostic Explanations (LIME) have been used with these models in order to identify the regions in the medical images that contribute towards the image classification of the cancer cells which can further provide a better understanding for performing diagnosis.