Periodic Reporting for period 2 - HYBRID (Innovative Training Network towards raising and supporting the next generation of creative and entrepreneurial cross-speciality imaging experts)
Reporting period: 2019-10-01 to 2021-11-30
HYBRID stands for “Healthcare yearns for bright researchers of imaging data”. We seek to provide essential information for personalized medicine that involves non-invasive imaging procedures. Personalized medicine is the practice of using selected and measured characteristics to guide decisions on prevention, diagnosis and treatment of disease for patient groups and individual patients. To realize the full potential of personalized medicine, highly-specific, reproducible and quantifiable biomarkers are required for disease characterization. Such biomarkers can be obtained with a number of techniques, of which non-invasive imaging methods have become a focus of attention over the past decade.
The ultimate goal of HYBRID is to establish the field of non-invasive disease characterization using various imaging methods. Specifically, we seek to adopt a more inclusive approach to image analysis, which considers anatomical and functional imaging using multiple of these methods, together with related non-imaging parameters that are extracted from patients during their clinical work-up. We propose this all-encompassing method, or holistic approach to deriving valuable information from imaging and non-image related data: HYBRID.
The HYBRID Mission
Our network brings together international academic, industrial and non-governmental organization partners in a cross-specialty platform to advance imaging sciences as part of personalized medicine with a focus on non-invasive imaging using multiple methods that includes anatomical and functional data from patients. Specifically, we present a consortium of world-leading imaging experts with in-depth knowledge of advanced imaging. All experts are embedded in a clinical environment, thus experienced in converting ideas and knowledge into applications fit for clinical use.
The HYBRID Vision
Our vision is to support the concept of personalized medicine by providing novel insights into disease characterization and diagnostic methods that are applicable in clinical routine.
Conclusions of the HYBRID project
The major scientific highlights of HYBRID include – but are not limited to - the derivation of harmonization standards for advanced PET/MR imaging and first approaches towards multi-centric radiomics evaluations (ESR1), the development of automated neural network-based segmentation of organs on PET (ESR7) and CT and MRI as a pre-requisite to SIRT therapy (ESR2), the development and validation of robust whole-body parametric maps as part of PET/MRI studies of humans (ESR5), as well as first advances towards the spatial alignments of histopathological sections and in-vivo imaging information in an effort towards deep learning models for classifying histopathology patches. In fact, many projects set out to embrace AI and machine learning to either automate data analytics or to deriving prediction models as part of clinical decision support, with the latter yielding a widely acknowledged contribution and peer-reviewed study on prostate cancer aggressiveness derived from PET/MR image data, resulting in improved cancer grading with respect to biopsy-based methods, with a potential clinical impact (ESR11).
Taken together, these examples attest to the successful conduct of the project, in that we propose a number of validated approaches to use biomarkers derived from dual-modality imaging systems (mainly PET/CT, PET/MR) for improved patient management.
The milestones and deliverables for the second reporting period have been completed. The progress of the projects has been presented at multiple (inter-)national conferences, and has so far resulted in 24 peer-reviewed papers.
HYBRID offered a wide range of training options for its fellows, including the following topics: scientific integrity, media training, (project) management skills, intellectual property rights, and open science. Moreover, annual training weeks have provided the fellows with a full overview of the techniques and methodological approaches available in HYBRID along with hands-on experiences. Concurrently with these annual training weeks, general meetings have been organized to encourage interaction between ESRs and PIs from the various beneficiaries.
The HYBRID website (www.hybrid2020.eu) provides an overview of the goals, mission and vision of the project and the (results of the) individual research activities. It is also the main place where news articles concerning HYBRID meetings, outreach activities by our fellows and other relevant information is shared. The articles are often (co)-written by fellows, who can use this as an opportunity to practice their writing skills for a general audience. Further dissemination of HYBRID news took place via the HYBRID Twitter account (@HYBRID2020). Finally, all fellows performed outreach activities every year to share their research with the general public, ranging from children to senior citizens.
It would be fair to say, that within HYBRID we managed to provide our ESRs with in-depth knowledge in medical physics, data processing and – to a lesser extent - biomedical engineering, all being united by a common understanding of the importance of “applied” science approaches, that is the usefulness and potential of translating methodological developments to the clinic. In that regard, we are confident that the readiness of ESRs post completion of HYBRID to work efficiently and effectively in medical sciences and medical industry is above average and beyond most state-of-the-art of singular university track education towards a PhD.
Ultimately, the results obtained by the HYBRID project will help groups of patients in the future by providing means to tailor diagnostic and therapeutic choices to individual patients. This, in turn, will bring healthcare costs down by avoiding futile and costly procedures for those who will not benefit from inappropriate and ineffective treatments.