Biomechanics in the Era of Personalised Health
In two months we will come together again for our annual meeting, incorporated into the World Congress of Biomechanics in Dublin. I am very much looking forward to this, especially as it gives us the opportunity to again give a European flavour to the WCB. It is also nearly two decades since the ESB held its meeting in Dublin, an event that for me coincided with my transition from doctoral candidate to “fully qualified” researcher, however one chooses to interpret that, with the opportunities and responsibilities to lead rather than follow.
Leading rather than following has been on my mind recently as I found myself in numerous discussions about personalised health, digitalisation and big data. At our own university and at the regional and national level in Switzerland, important new initiatives have been kicked off under the banner of personalised health, and I am fairly confident that this is the case across Europe. What I have found striking and somewhat disappointing in the presentations, the discussions and the strategies presented going forward is a heavy and quite uniform emphasis on the gene, the protein and the cell. Little attention has been paid to systems, interactions and functional responses occurring at a larger scale, or spanning from biology to other scientific domains. I think that we in the biomechanics community are in an excellent position, and indeed must not fail, to promote the inclusion of biomechanical insights into the framework of personalised health. Health is a basic and intrinsic thematic link through the majority of the work we do.
Comprehensive datasets covering whole populations are a central component of personalised health, and our community is in a unique position to be leaders, especially in the use of large-scale simulation to build from the bottom-up a profile of a whole population’s response. The critical need for personalised simulation is highlighted in a recently published evaluation of the merit of average biomechanical models, which demonstrated the short-comings of the traditional generic modelling paradigm. In contrast to traditional trials in the field of health sciences, in which a random patient population sample is subjected to individual experiments and subsequent statistical analyses, the generic modelling paradigm prescribes that the defining characteristics of a random sample of a target population are averaged beforethe input stage to a limited number of models. Unfortunately, due to the non-intuitive nature of non-linear systems, and the human body is a perfect example of such, average models provide results inconsistent with the average of individual tests/simulations on the population members.
In the position paper “in silicoClinical Trials: How Computer Simulation will Transform the Biomedical Industry”, with contributions from many of our ESB colleagues, a research and technical roadmapis presented for overcoming hurdles in the path of conventional development and validation of biomedical devices, and assessment of their efficacy, through the use of computational models.The roadmap provides compelling evidence for improvements in all stages of device design, pre-clinical verification and clinical assessment. The in silicoClinical Trial (ISCT) is defined in this document as “the use of personalised computer simulationin the development or regulatory evaluation of a medicinal product, medical device, or medical intervention. It is a subdomain of ‘in silico medicine’, the discipline that encompasses the use of individualised computer simulations in all aspects of prevention, diagnosis, prognostic assessment, and treatment of disease.” Indeed, the biomechanics community, through initiatives such as the Virtual Physiological Human, have been at the forefront of personalised health, but perhaps not always adequately recognised.
The biomechanics community is in a unique position to play an important role in personalised health. Mark Taylor and Patrick Prendergast highlighted in a recent critical reviewthe power (and limitations) of computational modelling, i.e. the ability to perform multiple analyses, varying individual input parameters across an appropriate range opening the door to comparative analyses, parametric studies, or even probabilistic assessments. However, true advances can only be achieved by modelling systems that are informed from robust data, underlining also the key contribution of physical measurements, capturing the relevant characteristics of an adequate population sample, and that incorporate the intricate interrelationships of the function of the human across multiple scales and across physical domains.
These are all challenges being addressed already in the scientific work presented at our annual meetings. We should strive to develop more strongly linked collaborative efforts to accelerate these advances, to share data towards comprehensive validation, and to encourage open discussion and comparison of “competing” efforts to capture the unique health-related response of an individual, against the backdrop of millions of individuals wishing to be equally served by translation of our science to improving human health. It’s a fascinating challenge, and I look forward to Dublin and gaining new insights into the unique biomechanical perspective on personalised health.
SF, Zurich, May 2018
Cook DD, Robertson DJ. The generic modeling fallacy: Average biomechanical models often produce non-average results! J Biomech. 2016 Nov 7;49(15):3609-3615.
Taylor M, Prendergast PJ. Four decades of finite element analysis of orthopaedic devices: where are we now and what are the opportunities? J Biomech. 2015 Mar 18;48(5):767-78.