UPF researchers create cutting-edge techniques to treat vertebral degeneration, one of the main causes of lower back pain

A European research consortium led by Pompeu Fabra University (UPF) has created innovative techniques to identify and treat spinal degeneration using artificial intelligence (AI) tools and computational techniques. These internationally pioneering techniques may provide fundamental clinical support to identify the causes of vertebral degeneration, one of the main causes of lower back pain, and design personalized treatments for each patient.

These are the results of the European project Disc4All, coordinated by Jérôme Noailly, head of the Biomechanics and Mechanobiology (BMMB) research area at the BCN MedTech Unit of the UPF Department of Engineering. The project, which began in November 2020 and has just been completed, has been supported by the European Commission through the Innovative Training Network (ITN) of the Marie Sklodowska-Curie Actions (MSCA), which back joint competitive research training programmes.

Noailly explains that the computational and AI models they have designed can contribute to more accurate diagnoses than those made with current clinical observations. “The complex interplay of factors is beyond our natural capacity of analysis. However, computer models and simulations can help us retain only the most important cause to consequence relationships, providing us with sufficient understanding for well-informed predictions and actions”, the Disc4All project coordinator, Jérôme Noailly, explains.

The project was driven by a European consortium, led by UPF and made up of research centres and technology and medical companies from eight countries, including experts in computer science and data, experimental and computational biology, bioinformatics, biomechanics, and medicine. In all, 15 doctoral projects have been coordinated that, together, have enabled this exceptional integration and fusion of disciplines. In addition to Noailly, other members of UPF who have been involved in supervisory tasks are Gemma Piella and Miguel Ángel González Ballester, also BCN MedTech principal investigators.

A project putting AI at the service of health

One of the main challenges of the research consortium was to design mathematical prediction models of intervertebral disc behaviour that reflect pathological biological changes and adjust to the diagnostic needs of clinicians and patients. “We began to further investigate the biological processes that actually trigger the degeneration of the spine. By gathering data on this, I thought we might be able to identify personalized and enhanced descriptors of the causes of pain”, Noailly explains. AI and machine learning allowed the research team to quickly integrate and process a large amount of data from diagnostic tests, laboratory experiments and computational analyses. In this way, objective risk factors for spinal degradation were identified along with demographic and psychological data that can also help predict the likelihood of suffering lower back pain. Processing these data brought about the creation of biomechanical and mathematical simulation models of the vertebrae.

The model simulations led to identifying potential biomarkers, that is, molecules that could be directly related to pain but have not been analysed clinically. Noailly managed to see patterns in biological processes that are not easy to capture in the traditional manner. “It’s a little like how astronomers describe faraway phenomena without actually seeing them. They use complex models, and this is what we are doing here”, Noailly explains. “AI-enhanced computer modelling is an efficient way of building up a personalized spinal model for precision medicine”, Noailly asserts. The goal is to transfer this pioneering computational modelling concept to the clinical field. For healthcare professionals, it can mean increased information with difficult-to-measure diagnostic factors, such as key biochemical changes in the disc according to its morphology, and better performance in the use of medical technologies such as magnetic resonance imaging.

An interdisciplinary consortium with partners from eight countries

The Disc4All project has been driven by a consortium of twelve beneficiaries and eight partner organizations. Apart from UPF, the beneficiary organizations are: InSilicoTrials (Italy), Barcelona Supercomputing Center-Centro Nacional de Supercomputación, Oulun Yliopisto (Finland), Galgo Medical SL (Spain), King’s College London (UK), Sheffield Hallam University (UK), Hospital del Mar Medical Research Institute (IMIM), University of Bern (Switzerland), ProATonce Ltd (Greece), University of Liège (Belgium) and Sheffield University (UK). The partner organizations are: Agency for Healthcare Quality and Evaluation of Catalonia, Technical Research Centre of Finland Ltd (Finland), Plexalis Ltd (UK), Virtual Physiological Human Institute (Belgium), QUAES Foundation (Spain), National Technical University of Athens (Greece), University of Barcelona (UB), and Rochester Institute of Technology (USA).

The project results highlighted on the EU’s Cordis platform

The results of Disc4All have been highlighted on the EU’s Cordis platform, dedicated to research and innovation, in the form of an article specifically dedicated to this topic and including it in a report on projects that stand out for applying AI to the field of life sciences.