The recent advent of high-throughput techniques for molecular analysis, including genomics, transcriptomics, proteomics, metabolomics, and imaging techniques, has allowed the storage of a large collection of data for identifying biomarkers used in the disease stratification, prediction, and early diagnosis of diseases [10, 11].

In this regard, the “radiomics” can be defined as the science that deals with the high-throughput extraction, storage, and analysis of a large amount of quantitative imaging features (imaging biomarker) in order to create accessible databases from radiological images and to reveal quantitative predictive or prognostic associations between images and medical outcomes [12, 13]. The modern and multiparametric imaging, characterized by digital and quantifiable informations, provides a set of biomarkers of the same patient that allow us to quantify the information. These biomarkers may refer to the organ function or neoplastic mass characteristics, and they are expressed by a number.

Examples of biomarkers are the diameter, volume, computed tomography (CT) density measurement, magnetic resonance (MR) signal intensity, standard uptake value (SUV) in positron emission tomography (PET) imaging, contrast enhancement (valuated in MR, CT, or ultrasound examination), perfusion parameters (i.e., blood flow, blood volume, mean transit time, and permeability), tissue elasticity in elastosonography, tissue pattern (texture analysis), morphological pattern, and much more. Moreover, beyond radiology, other types of images can be collected, for example, from endoscopy, microscopy, and surgery, providing measurable personalized data. Each of these biomarkers is patient specific and will be stored, analyzed, and correlated as part of a cluster of biomarkers of that patient [14].

The main focus of imaging biobanks is the “personalized medicine,” where the treatment is increasingly tailored on the basis of specific characteristics of the patient and their disease [15]. Quantitative medical imaging, with the identification of imaging biomarkers, represents a crucial part of personalized medicine providing selection criteria and follow-up strategies, tailored to the patient’s needs [16]. All these imaging informations should be considered as the phenotypic expression of a patient and can be correlated to the genotype. In this setting, the radiogenomics, which is the extension of radiomics, aims to identify a link between genotype and phenotype imaging [17].

The imaging biobanks are wide data collection including medical images and their correlated imaging biomarkers. The content of these biobanks, linked to that of biorepositories, should be available in a shared workflow among all researchers. A European network of imaging biobanks could significantly enhance the validation of new imaging biomarkers that could be potentially used as prognostic and predictive descriptors in the clinical practice.

In March 2014 the European Society of Radiology (ESR) instituted a dedicated working group (ESR WG on imaging biobanks) aimed at monitoring all existing imaging biobanks in Europe, promoting the federation and communication among them in a white paper [2]. Furthermore, the ESR Working Group promoted the realization of imaging biobanks and techniques for the analysis and processing of imaging biomarkers, stimulating the integration of existing image data repositories and also the link between the imaging biobanks and traditional biobanks, as well as encouraging the researcher cooperation for the standardization, validation and benchmarking of all data stored. The development of imaging biobanks is focused on imaging data collection and sharing for clinical research programs (i.e., clinical trials). By the definition and validation of new biomarkers, the imaging biobanks meet the need for storage, diffusion, and comparison of disease-specific data [14]. In this context, the international research collaboration promotes the comparing of imaging tools, protocols, data, and expertise, in order to establish common acquisition protocols and to ensure high image quality. These data collections could be based on the regional/national screening programs or clinical trials (i.e., performed for colorectal cancer, breast cancer, or lung cancer). The oncologic imaging represents the most suitable field for the discovery and validation of new biomarkers from multiple imaging modalities, since the oncologic patients are frequently monitored for staging and follow-up of treatment response. Nowadays, there is a significant need of detailed and accurate biomarkers, in order to reduce cancer morbidity and mortality, promoting the progression of the traditional “one size fits all” strategy toward a new “personalized” cancer therapy [18].

Accordingly to the dissemination and implementation of imaging biobank, the imaging collection and storage standardization are needed. The development of data standards promotes the communication among all the biobanks, using a standardized format, in order to integrate and share suitable informations for all researchers, as well as to provide legal regulation in the institutions [19]. Nowadays, this cooperation among all researchers about imaging biobanks is very poor. On these bases, the main focus of imaging biobanks is the endorsement of high-quality standard levels, yielding harmonized datasets for biomarker extraction thus reducing the inter-variability [20]. All imaging researchers should cooperate to improve and standardize the image acquisition protocols and archiving, the softwares for data analysis and processing, and further methodologies for imaging biobanks.