DeepCerviNet: Enhanced Cervical Cancer Classification Using Supervised Learning Model

Cervical cancer is a female-specific malignancy that poses a substantial threat worldwide, especially in less privileged and underdeveloped regions. Despite continuous progress and advances in medical science research, the rate of death from cervical cancer remains strikingly high among women. Therefore, accurate early-stage cervical cancer identification is imperative for effective diagnosis. It facilitates cost-effective and timely treatment to increase the survival rate of patients, leading to an optimal diagnosis improvement in disease classification. Deep learning approaches demonstrate superior accuracy in medical data classification, motivating us to consider these techniques in this research. In our study, we propose DeepCerviNet, a custom CNN model that achieves 99% precision, which exceeds the accuracy of other deep learning models, MobileNetV2 (97%), ResNET34 (95%), and GoogleNet (94%). Furthermore, to assess our model’s generalizability, we use another Multicancer dataset where the proposed model classifies various cancers with 99% accuracy. The DeepCerviNet model’s strong performance and adaptability can assist in identifying more aggressive or recurrent cervical cancer types, aiding the development of targeted therapeutic strategies.