Using AI to make society healthier

TNO works with commercial and health-related organisations to develop several platforms. These will provide lifestyle support, promote our understanding of our health and the necessary interventions and offer personalised advice. Our overall vision, is that AI should work together with citizens and healthcare providers to improve the individual health.

AI’s huge potential for personalising health

Personalised health, based on Artificial Intelligence, is an area in which we can manage preventive health measures, process data, develop predictive models and share that data between all parties in a secure environment. Healthcare nowadays is typically based on evidence from large-scale clinical trials. Health is personal, and so should be the advice. AI in healthcare, based on personal data from wearables/trackers and apps, consultation and medical data, offers new ways of making that advice personal. This can all be made possible through the personal health train (data secure infrastructure for data reuse and advice) and predictive models in health, in which the data of many citizens and patients can be connected and then fed to AI models.

Making AI understandable, explainable and controllable

There are two key challenges to the successful implementation of AI in healthcare. The first of these is making trustworthy and transparent algorithms in predictive health models. The second is managing the interaction between health professionals, individuals and the machine. Gaining consumers’ trust in any health advice that is aimed at improving their lifestyle is crucial. If they don’t trust that advice they simply won’t follow it. The outcomes of AI-enabled healthcare models must therefore be understandable, explainable, controllable and reliable for health professionals and individuals alike.

With respect to privacy, the data sources used in the creation of personalised health advice should be connected and the data must be shared in trustworthy, transparent and secure ways. Guaranteed confidentiality will be a key enabler in the development of new AI in healthcare applications.

Personalised Health: What does TNO offer?

• We develop privacy-by-design technology to safely share data.
• We combine biological, psychological and social data and insights with AI in healthcare techniques and expertise.
• We develop predictive models and related interventions.
• We develop digital biomarkers to diagnose health (including proxies).
• We design hybrid models (knowledge-data driven) and beta apps for diagnosis and health (behavioural change) (e.g. igrow).