This symposium focuses on the key approaches that yield detailed characterisation of brain disorders and modelling of this information for patient stratification and generating innovative treatment options.
Physiological, behavioural and environmental data collected via digital portable/wearable devices can provide a rich source of phenotypic information. In synergy with other “omics”, it promises to empower diagnostic interpretation and prediction. Personalized brain modelling can provide a basis for large-scale trials needed to strengthen the clinical evidence regarding such predictions. Identification of symptom groups based on fine-grained observations over large time spans and shared neural underlying may yield neurobehavioural constructs that can guide patient stratification and drug discovery. Human Brain Project workflows facilitate the registration of a wide range of data categories to reference schemes and atlases of the brain. This approach enables understanding the multi-scale organization of the brain and its pathologies.
Four pioneering projects combining multimodal data in large samples with computational approaches will be presented by the leading experts.
Artifical intelligence in clinical practice: where do we stand in 2019
Philippe Ryvlin, Lausanne (Switzerland)
Details anzeigen
Autor:in:
Philippe Ryvlin, Lausanne (Switzerland)
The development of artificial intelligence (AI) based methods in medicine has grown exponentially over the last decade, with several tens of thousands of scientific publications in the field. In contrast, a very limited number of clinical applications have been validated by regulatory bodies controling healthcare products. Indeed, the US Food and Drug Administration (FDA) has approved only a few AI-based technologies (less than 20) since 2017, while setting up a specific task force and novel fast-track process to facilitate such endorsement. The majority of FDA-approved AI-based solutions provide diagnostic tools for more effective interpretation of medical images. More sophisticated and multidimensional approaches towards precision medicine and enhanced patients’ management remain hindered by the global challenge of accessing adequately organized clinical data, an issue tackled by the medical informatic platform of the Human Brain Project (HBP). Other significant issues in AI-based medicine include the principles of transparency, safety and accountability, reimbursement, and AI-literacy of healthcare providers.
Modelling of brain activity and connectome in pathological states for therapeutic targeting
Viktor Jirsa, Marseille (France)
Details anzeigen
Autor:in:
Viktor Jirsa, Marseille (France)
Pattern formation in physics, biology and chemistry is based on dynamic principles of self-organization. Pattern formation phenomena in brain networks are no exception and form the basis of our current understanding of cognitive brain functions. Perception and motor behavior emerge together with neuroelectrical and chemical activity, but so do pathological disorders. This pattern formation in the brain results from the interaction of billions of neurons over several time and space scales, but is typically measured in humans only on very large scales such as in magnetic resonance imaging or electroencephalography (EEG). In order to bridge the gap to clinical applications, it is therefore essential to model the traverse of scales using computer simulations and advanced mathematics, supported by individual state-of-the-art brain imaging. This combination of approaches has been developed within The Virtual Brain consortium and nowadays allows to create autonomous brain models of individual patients and to test concrete clinical questions, possibly even to develop new therapies. Especially in epilepsy, these modern approaches have been successfully applied to the development of novel surgical interventions and network modulation, paving the road for future therapeutic targeting of pathological states.