A new era in personalised medicine: Radiomics as decision support tool for diagnostics and theragnostics in oncology

The high degree of tumour (genomic and phenotypic) heterogeneity influences patient’s response to therapy and hampers wide deployment of personalised medicine for cancer treatment. Thus, there is an imperative need for new technologies that can accurately detect tumour heterogeneity, allow for patient stratification and assist clinicians in providing the right diagnosis and treatment for the right patient. PREDICT’s mission is to address this huge unmet need. Radiomics, a newly emerging field that uses high-throughput extraction of large amounts of features from radiographic images, can boost the field of personalised medicine. The analysis of medical images taken as standard-of-care allows Radiomics to capture tumour heterogeneity and to generate ‘tumour-specific’ signatures in a non-invasive way, without the need of assessing the patient’s genetic profile. Thus, Radiomics, if linked to Big- data and decision support systems (DSS), can be used as diagnostic tool for patient stratification, for prediction of treatment response and for guidance, involving the
patient, of clinical decisions in oncology. However, researchers that understand cancer biology, advanced imaging and big data analytics are virtually absent. Even more challenging is to translate the outcomes into actual clinical tools involving the patient.
PREDICT builds upon a strong consortium with 8 academic and 10 non-academic partners that are all pioneers in their respective field.

The impact of the PREDICT network

PREDICT will train 15 highly promising young investigators to become professionals and business experts in Radiomics, Big Data, machine learning approaches, and multifactorial DSS. These fields that are currently considered the future in diagnostics and theragnostics for different types of cancer and non-cancerous chronic diseases. PREDICT will provide a systematic and multidisciplinary curriculum of courses and research experience so these ESRs will be able to address the current bottlenecks in cancer diagnostics and theragnostics. But most importantly these ESRs to be the professionals to drive innovation and apply their expertise in other areas as well (e.g. further expand the application of Radiomics in other types of cancer, apply the lessons learned from medical imaging analysis/processing and machine learning approaches to other diseases areas). A direct consequence of Radiomics in diagnostics and theragnostics will be the reduction of side-effects related to sub-optimal treatment of cancer patients and reduction of costs associated with drug failure, while ensuring rational use of the limited healthcare resources. PREDICT will facilitate personalised medicine for cancer patients and as such will:

1. provide ‘tumour-specific’ signatures for cancer diagnostics and theragnostics that will guide clinical decisions and enable the right treatment for the right person,

2.  advance the field of participative medicine,

3. provide solid evidence for proper stratification of patients in clinical trials, thus minimising the risk of drug failure, and

4. stimulate biotech and big pharmaceutical companies to develop more targeted therapies and to use big-data Radiomics platforms.

Furthermore, lessons learned from PREDICT concerning big data analytics, machine learning approaches, distributed learning and multifactorial DSS can further be applied in other areas such as chronic diseases (Alzheimer, COPD), cardiology and any medical field based on imaging. Within the project, the consortium aims to create a robust, but flexible and fully automated Radiomics platform for use in cancer diagnostics and theragnostics to provide clinical proof-of -concept, which will fuel industrial interest. It is therefore evident that the benefit of PREDICT will become evident at multiple levels: scientific, clinical, technological innovation and societal.

The partners in the PREDICT network

Maastricht University | Netherlands
Fondazione IRCCS Istituto Nazionale Dei Tumori | Italy
Universite de Liege | Belgium
Deutches Krebs Forchungs Zentrum | Germany
Institut National de la Sante et de la Recherche Medicale | France
Oncoradiomics | Belgium
Mirada Medical | United Kingdom
Perspectum Diagnostics | United Kingdom
Health Innovation Ventures | Netherlands

Associated partners:
DualTPharma | Netherlands
Elsevier | Netherlands
European Cancer PAtient Coalition | Belgium
Siemens | Germany
Universita Cattolica S. Cuore | Italy
Heidelberg University Medical School | Germany