Development of a preoperative predictive model of malignancy of indeterminate ovarian masses

Abstract

Technological innovations have transformed healthcare, offering new avenues for disease diagnosis and treatment. Among these advancements, radiomics analysis of MRI images holds great promise for improving the early detection and characterization of ovarian masses. Ovarian cancer, with its high mortality rates attributed to late-stage diagnosis, stands to benefit significantly from these developments. This study focuses on exploring various pathways before applying radiomics analysis to ovarian masses in MRI images. By investigating different preprocessing methods, such as intensity normalization and registration of MRI images, the project aims to determine their impact on the evaluation of radiomic features. Through a comparative analysis of the radiomic feature results obtained from these pathways, this research seeks to understand how variations in preprocessing approaches affect the evaluation of radiomic features. Despite challenges such as limited patient samples and time constraints, the study anticipates significant outcomes. It aims to provide understanding about how various preprocessing techniques impact the assessment of radiomic features. Additionally, it explores the potential for developing a specialized pipeline that incorporates the most effective preprocessing methods for practical use in clinical environments. By advancing the field of ovarian cancer diagnosis through radiomics and MRI analysis, this research has the potential to improve patient outcomes and healthcare practices. Ultimately, the goal is to contribute to early detection and personalized treatment strategies, ultimately reducing the burden of ovarian cancer on patients and healthcare systems.