•  
  •  
 

Abstract

Radiomics is the study of features extracted from different radiology related imaging modalities. The objective of radiomics is to enhance the ease and quality of diagnosis. In particular, radiomics information are integrated with the clinical findings to arrive at diagnostic prediction which otherwise may involve further intensive clinical and pathological examination. For example, image features extracted from PET-CT when combined with clinical and pathological information, may be used as an early test for malignancy even before gold standard biopsy test. This review discusses the use of radiomics for intensity, statistical and textural features of non-invasive imaging modalities. Example modalities include images from Positron Emission Tomography (PET), Computed Tomography (CT) and Magnetic Resonance Imaging (MRI). There are different publicly available packages like PyRadiomics, LIFEx for extraction of radiomic features. While reporting results using radiomics, there are wide variations in results. This necessitates standardization of features and standardization of quality measures. The Image Biomarker Standardization Initiative (IBSI) takes care of standardization of radiomics features while standardization of quality measures is maintained through Radiomics Quality Score (RQS). Both IBSI and RQS are discussed in this review. To follow a systematic process of review, Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) is also followed.

Share

COinS