An Exploratory study for characterIzing and predicting prostate abNormalities uSing MRI-based radiomics and arTificial intEllIgeNce (EINSTEIN)

Introduction Prostate cancer (PCa) is the fourth most prevalent cancer globally, and the most common among men. Most PCa patients in Oman are presented during the advanced stages of the disease with widespread metastatic disease reducing their overall rates of survival. Characterisation of the Omanis PCa population could be beneficial to develop a clinical profile demonstrating specific characteristics to better classify and derive radiomics signatures. These could help in developing artificial intelligence methods to assist with earlier and quicker diagnosis of possible prostate lesions. Methods A retrospective, cross-sectional study has been designed to determine the pathological and radiological characteristics based on multi-sequence 3-dimensional Magnetic Resonance Imaging (MRI). The MRI records are maintained within the existing electronic healthcare records of the Sultan Qaboos University Hospital’s Department of Radiology and Molecular Imaging. Data will be extracted based on a confirmed diagnosis reported between the 1 st of January 2010 and October 2023. All patients included within the study will be aged between 18-99 years. A study specific data extraction template has been devised to gather demographic details, clinical parameters and radiological findings based on existing imaging reports within the HIS and PACS systems. Ethics approval Research Ethics approval reference for this study is (MREC #3176 REF. NO. SQU-EC/ 283\2023) Conclusion The data analysis will be conducted using statistical software to conduct a Joinpoint regression analysis and linear regression modelling. We will also conduct a descriptive analysis.