RADIOMICS

RADIOMICS

Radiological images contain textural information that is not visually perceivable to radiologists and which could potentially decode tissue pathology. Our approach is built on the concept of radiomics and deals with development of full textural images called the radiomic feature maps from radiological images. The radiomic feature maps provide excellent textural visualization of breast multiparametric MRI in addition to correlation with tissue biology of both normal and cancerous tissue. Furthermore, the radiomic feature maps obtained from different MRI parameters were combined into an automated classification model to identify a breast lesion as benign or malignant with very high accuracy. The radiomic feature maps hold the potential for development of non-invasive quantitative biomarkers for personalized clinical diagnosis and treatment planning.

Publications

  • V. Parekh and M. A. Jacobs, “Radiomics: a new application from established techniques.” Expert Review of Precision
    Medicine and Drug Development. 2016 Mar 3;1(2):207-26.
  • V. S. Parekh and M. A. Jacobs, "Radiomic feature mapping framework for breast cancer diagnosis and correlation with
    breast cancer biology," Submitted to Nature Breast Cancer, 2017.