THE USE OF ARTIFICIAL INTELLIGENCE ON COLPOSCOPY IMAGES AND SEGMENTAL VOLUMES, CONSTRUCTEDFROM MRI AND CT IMAGES, IN THE DIAGNOSIS AND STAGING OF PRECANCERS, CERVICAL CANCERS ANDTHYROID CANCERS

Aim of the study For the accurate diagnosis and staging of precancers and cervical and thyroid cancer, we aim to create a diagnostic method optimized by artificial intelligence (AI) algorithms and validated by the notable positive results of a randomized, controlled, 17-month trial.. Materials and methods The optimization of the method will involve the development and training of artificial intelligence models using convolutive neural networks (CNN) to identify precancers and cancers in colposcopic images. We will use topologies that have demonstrated strong performance in similar image recognition projects, such as VGG16, Inception, MobilNet, ROI, U-Net, and KiU-Net. Additionally, the research includes a comparative study of various algorithms and tools employed in segmental volumetric constructions to generate 3D images from MRI/CT data. This study will also evaluate current advancements in DICOM image processing using techniques such as volume rendering, transfer functions for opacity and color, and shades of gray from DICOM images projected in a three-dimensional space. Validation of the proposed method will be achieved through a randomized, controlled clinical trial conducted over 17 months. Patients will be informed and recruited either via random presentation at the specialized medical centers participating in the trial or through a dedicated web platform. Selection criteria will adhere to inclusion and exclusion parameters defined in the clinical trial protocol, and all patient data will be handled ethically and in accordance with written and informed consent approved by the Ethics Committee. . Results The optimized method, supported by AI algorithms and validated through clinical trials, aims to demonstrate concrete and favorable outcomes in the diagnosis, staging, and treatment planning for cervical and thyroid precancers and cancers.. Conclusions By implementing this AI-optimized diagnostic method, we seek to raise the quality standard in diagnosing and staging precancers and cancers, ultimately enhancing therapeutic decision-making and patient outcomes.