Enhancing cervical cancer detection and robust classification through a fusion of deep learning models

Area of Science:

• Oncology
• Medical Imaging
• Artificial Intelligence

Background:

• Cervical cancer is a leading cause of cancer death in women globally.
• Current screening methods like Pap smears and HPV testing require expert interpretation, highlighting the need for improved diagnostic tools.
• Early detection is critical for effective cervical cancer treatment and improved patient outcomes.

Purpose of the Study:

• To develop and evaluate a novel hybrid deep learning (DL) and machine learning (ML) methodology for automated cervical cancer classification.
• To leverage pre-trained deep neural networks for sophisticated feature extraction from cervical images.
• To assess the performance of various ML algorithms integrated with DL features for enhanced diagnostic accuracy.

Main Methods:

• Utilized pre-trained deep neural network models (Alexnet, Resnet-101, Resnet-152, InceptionV3) for image feature extraction.
• Fine-tuned DL models and integrated diverse ML algorithms for cervical cancer classification.
• Employed the publicly accessible SIPaKMeD dataset for model training and validation.

Main Results:

• The hybrid DL-ML methodology demonstrated high efficacy in cervical cancer detection.
• ResNet152, as part of the hybrid model, achieved a remarkable accuracy of 98.08%.
• The approach successfully extracted complex features from medical images, facilitating accurate classification.

Conclusions:

• The proposed AI-driven methodology offers a promising advancement in automated cervical cancer screening and diagnosis.
• Intelligent automation through hybrid DL-ML models can significantly improve the accuracy and efficiency of medical diagnostics.
• This approach paves the way for more timely and precise interventions in cervical cancer management.