A simulation and prediction tool for paediatric clinicians
Has the growing political and scientific focus on active and healthy ageing come at the expense of children’s health? Prof. Bruno Dallapiccola, OPBG Scientific Director and Project Coordinator, certainly thinks so. With a consortium of 22 organisations from across Europe, he — with the support of Prof Edwin Morley Fletcher, president of Lynkeus and Project Manager — has spent the past four years developing an advanced digital repository (infostructure) that integrates data from the likes of clinical, genetic and metagenomic analysis, MRI and US image analytics, haemodynamics, real-time processing of musculoskeletal parameters, and fibres biomechanical data. The platform allows clinicians to look for similarities with their own patients, access model-based simulations and predictions, and look for patient-centric clinical workflows. ‘Our idea stemmed from the need of optimizing existing treatments towards a predictive and personalised clinical approach, allowing to foresee the outcomes of clinical interventions and to tailor them to a single patient’s physiological parameters. The ultimate goal is to reduce medical errors and suboptimal treatments, as well as decrease overall medical costs,’ Prof Dallapiccola explains. In addition to the fact that ‘our world is becoming increasingly unfair towards the younger generations’, he observes that diseases in infants, children and adolescents are a large and underestimated public health problem. One that the MD PAEDIGREE (Model-Driven European Paediatric Digital Repository) project is meant to help overcome. MD PAEDIGREE hosts data leveraged by advanced analytics tools such as deep machine learning or similarity search, so as to identify hidden common patterns. From thereon, physicians can build personalised models able to reproduce the individual patient’s physiological parameters — either at pre-interventional level or as result of a given clinical intervention — and categorise patients based on disease risk. ‘Using these tools would minimize the chance of medical error and increase treatment efficacy, reducing in turn the risks of complications and relapse, time of recovery and clinical costs,’ Prof Morley-Fletcher points out. A promising future Prof Dallapiccola is proud of the project outcomes, and says that early reactions from clinicians involved in the project have been noticeably welcoming and encouraging. ‘Although the user interface has not yet reached the maturity level required for a seamless integration in everyday clinical practice, clinicians have largely recognised the added value of the implemented technological solutions, particularly for supporting their clinical decision making,’ he explains. ‘Also, the tentative model integration in the clinical workflow has been highly appreciated, allowing for intense engagement of clinical partners at every step of the model-driven workflow implementation.’ Although the project was completed at the end of February, the MD-Paedigree repository will continue to be fed with routine clinical datasets coming from the affiliated clinical centres, and various research initiatives are already building upon its results. A project named CARDIOPROOF adopted the very same infostructure for conducting its research on model-based cardiology prediction, and the MHMD project has been implementing an innovative EU-based platform for sharing of patients’ electronic health records (EHR) by including and extending the MD-Paedigree infostructure. ‘I am confident that several research initiatives will continue to flourish upon the project outcomes, as MD-Paedigree has provided tools and systems which will contribute to pioneering the new era of precision medicine,’ Prof Dallapiccola concludes.
Keywords
digital repository, paediatric clinical data, infostructure, MD PAEDIGREE, patient, treatment, clinical datasets, CARDIOPROOF, MHMD, electronic health records
