Radiomics in [¹⁸F]FDG PET/CT: A leap in the dark?

Positron emission tomography (PET) imaging with the non-metabolisable glucose analogue 2-[¹⁸F]fluoro-2-deoxy-D-glucose ([¹⁸F]FDG), combined with low dose computed tomography (CT) for anatomical reference, is an important tool to detect and stage cancer or active inflammations. Visual interpretation of PET/CT images consists of (qualitative) assessment of radiotracer uptake in different tissues and their density. Furthermore, the location, size, shape, and relation with surrounding tissues of these lesions provide important clues on their nature. Yet, medical images contain much more information about tissue biology hidden in the myriad of voxels of both lesions and healthy tissue than can be assessed visually. Quantification of radiotracer uptake heterogeneity and other tissue characteristics is studied in the field of radiomics. Radiomics is a form of medical image processing that aims to find stable and clinically relevant image-derived biomarkers for lesion characterisation, prognostic stratification, and response prediction, thereby contributing to precision medicine. Radiomics consists of the conversion of (parts of) medical images into a high-dimensional set of quantitative features and the subsequent mining of this dataset for potential information useful for the quantification or monitoring of tumour or disease characteristics in clinical practice.

This thesis contributed to a deeper understanding of the methodological aspects of handcrafted radiomics in [¹⁸F]FDG PET/CT, specifically in small datasets. However, most radiomic papers present proof-of-concept studies and clinical implementation is still far away. At some point in the future, radiomic biomarkers may be used in clinical practice, but at the moment we should acknowledge the limitations of the field and try to overcome these. Only then, we will be able to cross the translational gap towards clinical readiness. Future research should focus on standardisation of feature selection, model building, and ideally a tool that implements these aspects. In such a way, radiomics may redeem the promise of bringing forth imaging biomarkers that contribute to precision medicine.