Closing date 4 February 2020
D-Risc: Multiscale modelling & data mining for intervertebral disc degeneration risk prediction
D-Risc aims to reveal critical interplays of crucial stimuli within the intervertebral disc (IVD) that might lead to IVD degeneration, based on morphological and physiological parameters. Different models at the organ, tissue and cellular levels will be used. Specifically, the project will combine multi-physics finite element models at the organ and tissue levels with agent-based and network models at the cell and molecular levels, to simulate the local regulation of IVD cells in multifactorial physical and biochemical micro-environments. Simulation results will be mined with patient-specific morphological, physical activity and life-style data. Depending on the identified multiscale paths that can lead to degeneration-related cell activity (i.e. catabolic shift of cell activity), personalised recommendations for prevention- and optimised conservative treatments will be established.
D-Risc will exploit the competencies of the Biomechanics and Mechanobiology (BMMB - http://biomech.es) lab of the BCN MedTech research unit at the Department of Information and Communication Technologies (DTIC) of the Universitat Pompeu Fabra (UPF), Barcelona. The project will be additionally implemented in cooperation with the medical image analysis and machine learning areas of BCN MedTech (http://bcn-medtech.upf.edu/).
UPF was established in 1990 as a public university with strong dedication to excellence in research and teaching. It is the 1st Spanish university in the world Top 200 (THE2020), the 11th (ranked 5th in Europe and 1st in Spain) under 50 years (THE18). It is the Spanish university department with the largest number of ERC grants (9 FP7 and 9 H2020) and is part of the FET Flagship initiative “The Human Brain Project”.
BCN MedTech is the Barcelona Centre for New Medical Technologies at UPF. It focusses on biomedical integrative research, including mathematical and computational models, algorithms and systems for computer-aided diagnosis and treatment, and the translation thereof into relevant clinical problems and industrial products. It has a team of 60 full time researchers working on medical image and signal processing, computational simulation, computer-assisted surgery and biomedical electronics. Within BCN MedTech, the BMMB lab combines mechanistic and stochastic theoretical modelling with computational methods in biology and physics, to rationally explore the complex multiscale interactions between tissue multiphysics and biological processes, and to understand the bottom-up regulation of the functional biomechanics of organs in health and disease. The specific targets are cartilaginous (rheumatic disorders), bone (osteoporosis), arterial (atherosclerosis) and lung (emphysema) tissues. The project will combine this expertise with computational anatomy and manifold learning techniques for patient.
D-Risc capitalizes on previous research at BCN MedTech, to assess the risk of disc degeneration (DD). Patient-specific IVD finite element (FE) models will be coupled to agent-based (AB)/network cell models, to predict catabolic shifts of cell activity in function of morphological, metabolic and mechanical factors. According to subsequent machine learning analyses, specific combination of factors will be identified as possible risks for DD.
Low back pain (LBP) affects up to 85% of people at some point in life. It is strongly related to DD, with phenotypes that cannot be explained solely by genetic factors as they also depend on mechanical loads.
In vivo or in-vitro studies investigated DD at the cell and tissue levels, but they are costly and limited in terms of parameterization, effective number of measurements and long-term observations. In contrast, computational modelling allows testing different boundary conditions (mechanical, biochemical, …) and numerous theoretical hypotheses over long timescales, at a relatively limited cost.
Coupled to personalized organ models, multiscale models and simulations can indicate common patterns in specific groups of IVD, as well as critical combinations of cell stimuli and the effects thereof on DD observable features. In particular, the mining of model inputs together with simulated data can reveal such patterns and combinations.
The successful candidate will join the BCN MedTech team and will be co-supervised by faculties, experts in computational multiscale modelling and machine learning. (S)He will systematically analyze the 3D anatomy of 500 patient-specific IVD FE models, available at UPF, to define relevant groups of FE /AB multiscale simulations. Then, (s)he will use machine learning algorithms to build correlation models among personalized model inputs and predicted cell activity, to create a DD risk score model.
D-Risc will involve key collaborations with population cohort infrastructures in UK and Finland.
If you're interested to apply full information can be found here