Affiliations 

  • 1 Division of Imaging Sciences and Technology, School of Medicine, Ninewells Hospital, University of Dundee, Dundee DD1 9SY, UK
  • 2 Tayside Centre for Genomic Analysis, School of Medicine, University of Dundee, Dundee DD1 9SY, UK
  • 3 Department of Pathology, Ninewells Hospital, Dundee DD9 1SY, UK
  • 4 Surgical Skills Centre, Dundee Institute for Healthcare Simulation Respiratory Medicine and Gastroenterology, School of Medicine, Ninewells Hospital and Medical School, University of Dundee, Dundee DD1 9SY, UK
  • 5 Division of Population Pharmacogenetics, Population Health and Genomics, Biomedical Research Centre, Ninewells Hospital and Medical School, University of Dundee, Dundee DD1 9SY, UK
Int J Mol Sci, 2024 Nov 21;25(23).
PMID: 39684226 DOI: 10.3390/ijms252312512

Abstract

RO and ChRCC are kidney tumours with overlapping characteristics, making differentiation between them challenging. The objective of this research is to create a radiogenomics map by correlating radiomic features to molecular phenotypes in ChRCC and RO, using resection as the gold standard. Fourteen patients (6 RO and 8 ChRCC) were included in the prospective study. A total of 1,875 radiomic features were extracted from CT scans, alongside 632 cytobands containing 16,303 genes from the genomic data. Feature selection algorithms applied to the radiomic features resulted in 13 key features. From the genomic data, 24 cytobands highly correlated with histology were selected and cross-correlated with the radiomic features. The analysis identified four radiomic features that were strongly associated with seven genomic features. These findings demonstrate the potential of integrating radiomic and genomic data to enhance the differential diagnosis of RO and ChRCC, paving the way for more precise and non-invasive diagnostic tools in clinical practice.

* Title and MeSH Headings from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.