RADIOGENOMICS IN BREAST CANCER: INTEGRATION OF RADIOLOGICAL AND GENOMIC DATA

Abstract

Article History

Received:

July 13, 2023

Revised:

August 18, 2023

Accepted:

September 07, 2023

 

 A powerful paradigm of breast cancer personalized treatment is presented by the combination of genetic profiling and medical imaging, namely, radiogenomics.  This study presents a radiogenomic modeling mixed-methods experimental examination of digital mammography and high-resolution MRI assortments of patients with verified histopathologically confirmed breast cancer.  Radiomic features like enhancement pattern, shape, and margin of a tumor were extracted across standardized computation pipelines. These characteristics were subsequently cross-referenced with genomic profiles which represented somatic mutations, gene expression patterns and hormone receptor status.  Multivariate regression and machine learning models were fitted to predict the clinical outcomes including categorization of molecular subtypes and therapeutical responsiveness.  It has an overall AUC of 0.87 in predicting subtype classification indicating a high level of predictive ability in the integrated model.  The unsupervised clustering has established different radiogenomic symptoms with prognostic implications.  The qualitative comments presented by oncology experts confirmed the clinical interpretability of the proposed methodology and the possibility of integrating them into the workflow.  Some of the ethical concerns that were addressed included handling of genomic data, explaining the model and acceptance into clinical practice.  The ongoing cycle between imaging acquisition, image processing, genomic profiling, multimodal information integration, and their association with clinical information is proved by the end-to-end process, which is shown in Fig. 1.  All things being put into consideration, this study advocates the feasibility and potentiality of radiogenomics in precision treatment of breast cancer.