Development and Evaluation of an Artificial Intelligence Model for Cervical Cancer Detection From Colposcopic Images

What this Trial Is About

Cervical cancer remains a serious health challenge, especially in low-income and middle-income countries where late diagnosis and limited access to screening contribute to high death rates. This research is focused on creating and testing an artificial intelligence (AI) model that can analyze images taken during colposcopy, a procedure used to look closely at the cervix for signs of cancer. The goal is to improve the accuracy and efficiency of detecting cervical abnormalities, reducing dependence on doctors' experience, which can vary, and overcoming the limitations of invasive biopsies. The study involves both collecting new and reviewing past colposcopic images and related clinical information from women undergoing screening at a medical center in Bangladesh. The AI model will be developed using this diverse set of images to recognize transformation zones and detect cancer stages. Colposcopy involves applying saline, acetic acid, and iodine to the cervix during a single session lasting about 10 to 15 minutes. The study group includes women with suspected cervical abnormalities and those without, to build a comprehensive dataset and test the model's generalizability. Participants will undergo colposcopic examinations, and their images and medical data will be collected for AI analysis. The study will assess the model's accuracy using measures like the Swede Score and its ability to classify transformation zones. Quality assurance, data validation, and statistical analysis plans are in place to ensure reliable results. The research aims to evaluate how well the AI tool supports clinical decisions and improves early detection, with the potential to make cervical cancer screening more accessible and effective in resource-limited settings.

AI Model for Cervical Cancer Detection From Colposcopy Images