Tumour Phenotype with non-invasive imaging

Welcome to Radiomics. Please view the animation below.

About Radiomics

Human cancers exhibit strong intra and inter patient heterogeneity, which occurs at different levels: genes, proteins, cells, microenvironment, tissues and organs. This limits the use of, for instance, biopsy based molecular assays but in contrast gives a huge potential for non-invasive imaging techniques. Over the past decade, the use and role of medical imaging technologies in clinical oncology has greatly expanded from primarily a diagnostic tool to include a more central role in the context of individualized medicine (figure 1). Recent advances in medical imaging technology allow the use of more advanced image analysis methods beyond simple measurements of, for instance, tumor size or radiotracer uptake metrics. Imaging therefore has great potential to guide therapy and to monitor the development and progression of the disease or its response to therapy.

Radiomics enables the high-throughput extraction of a large amount (400+) quantitative features from medical images of a given modality (e.g. CT, PET, or MR), providing a comprehensive quantification of the tumor phenotype, based on simple medical imaging. Radiomics can provide complementary and interchangeable information compared to other sources (e.g. demographics, pathology, blood biomarkers, or genomics), improving individualized treatment selection and monitoring. Radiomics can have large clinical impact, since imaging is routinely used in clinical practice worldwide, providing an unprecedented opportunity to improve decision-support at low cost.

Radiomics: workflow

The Radiomics workflow basically consists the following steps (Figure 3). The first step is acquisition of high quality standardized imaging, for diagnostic or planning purposes. The macroscopic tumor is defined on these images, either with an automated segmentation method or alternatively by an experienced radiologist or radiation oncologist. Next, a large number of quantitative features is extracted from the previously defined tumor region. These features describe, amongst others, tumor image intensity, texture and shape and size of the tumor. The final step is analysis of the acquired imaging features.

Figure 1: Different sources of information, e.g. demographics, imaging, pathology, toxicity, biomarkers, genomics and proteomics, can be used for selecting the optimal treatment.

Figure 2: Typical examples of lung tumors, showing differences captured by simple CT imaging.. Displayed regions are 16x16 cm for the displayed CT slices and 16x16x16 cm for the 3D renderings.

Figure 3: Schematic representation of the workflow

Radiomics Digital Phantom

The field of radiomics is emerging rapidly; however, the field lacks standardized evaluation of both the scientific integrity and the clinical significance of the numerous published radiomics investigations resulting from this growth. There is a clear and present need for rigorous evaluation criteria and reporting guidelines in order for radiomics to mature as a discipline. We therefore provide guidance together with a novel metric, the radiomics quality score (RQS) and an online digital phantom (DOI:10.17195/candat.2016.08.1), to meet this urgent need for both past and future investigations in the field of radiomics. The objective of the Radiomics digital phantom is to compare different software implementations for radiomic feature extraction algorithms, we provide CT data of the primary tumor region and the corresponding tumor contours of four lung cancer cases, to serve as “real life” digital phantoms. Using the pre-processed image data, we calculated a set of commonly used features to serve as a reference feature dataset. Please see the supplementary material (DOI:10.17195/candat.2016.08.1), for a detailed description of the digital phantom image data and calculated features.

List of scientific publications

Parmar C, Rios Velazquez E, Leijenaar R, et al. Robust radiomics feature quantification using semiautomatic volumetric segmentation. PLoS One. 2014;9(7):e102107.

Hugo J. W. L. Aerts, Emmanuel Rios Velazquez, Ralph T. H. Leijenaar, Chintan Parmar, Patrick Grossmann, Sara Cavalho, et al. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat. Commun. 2014;5: 4006.

Lambin P, Rios-Velazquez E, Leijenaar R, Carvalho S, van Stiphout RG, Granton P, et al. Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer 2012;48: 441-446.

Velazquez ER, Parmar C, Jermoumi M, Mak RH, van Baardwijk A, Fennessy FM, et al. Volumetric CT-based segmentation of NSCLC using 3D-Slicer. Scientific reports 2013;3: 3529.

Leijenaar RT, Carvalho S, Velazquez ER, van Elmpt WJ, Parmar C, Hoekstra OS, et al. Stability of FDG-PET Radiomics features: an integrated analysis of test-retest and inter-observer variability. Acta Oncol 2013;52: 1391-1397.

Lambin P, van Stiphout RG, Starmans MH, Rios-Velazquez E, Nalbantov G, Aerts HJ, et al. Predicting outcomes in radiation oncology--multifactorial decision support systems. Nat Rev Clin Oncol 2013;10: 27-40.

Lambin P, Roelofs E, Reymen B, Velazquez ER, Buijsen J, Zegers CM, et al. 'Rapid Learning health care in oncology' - An approach towards decision support systems enabling customised radiotherapy'. Radiother Oncol 2013;109: 159-164.

Carvalho S, Leijenaar RT, Velazquez ER, Oberije C, Parmar C, van Elmpt W, et al. Prognostic value of metabolic metrics extracted from baseline positron emission tomography images in non-small cell lung cancer. Acta Oncol 2013;52: 1398-1404.

Rios Velazquez E, Aerts HJ, Gu Y, Goldgof DB, De Ruysscher D, Dekker A, et al. A semiautomatic CT-based ensemble segmentation of lung tumors: comparison with oncologists' delineations and with the surgical specimen. Radiother Oncol 2012;105: 167-173.