Frequently asked questions

Where is the name “Radiomics” coming from?

Radi refers to Radiology. The suffix Omics means a study of the totality of something; e.g. genomics, proteomics. In the case it refers to the totality of radiological images. Radiomics aims at the collective characterization and quantification of pools of image features. In short radiomics is a synonymous of “Quantitative Imaging”

The basic information (the underlying image) is fixed and will not change. Why/how do you get more information with Radiomics?

•We provide quantitative information (objective), that goes beyond simple measurements.
•It’s high throughput. Like a facial recognition software analyzing a crowd.
•Because Radiomics software can see information which is not seen and/or quantifiable by the human eye.

Is Deep Learning part of Radiomics?

Yes there are two families of Radiomics methods: handcrafted Radiomics with predefined features and Deep Learning using  Convolutional Neural Networks.

Is tumour volume a radiomics feature?

Yes, it is. Tumour volume (IBSI: “original_shape_VoxelVolume”) is a quantifiable imaging variable and is therefore a radiomics feature. Tumour volume is surprinsingly rarely used in clinical routine. RECIST and TNM tend to use the largest diameter of tumours.

Is Radiomics limited to Oncology?

No it is applicable to any medical images in any field.

How and when was Radiomics invented?

The idea of Radiomics was born in the lab of Philippe Lambin in 2008 in Maastricht after several formal and informal discussions with Bob Gillies during his sabbatical as TEFAF Chair Recipient in Maastricht. At that time Hugo Aerts was still PhD student in Philippe’s lab before he moved to Harvard.

When was the name Radiomics used for the first time in a scientific publication?

In Pubmed the first publications referring to “Radiomics” was the following: Lambin P et al. HJ Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer. 2012 Mar;48(4):441-6.