Providing the right diagnosis as early as possible, giving the most appropriate treatment and predicting the evolution of the disease for a specific patient is of paramount importance in precision medicine. However, diseases are dynamic and very heterogeneous. Their etiologies are very complex and often several hypotheses are needed to explain their pathogenesis. Most of the time it is experimentally very difficult to understand how the interactions of the various pathways and mechanisms lead to the pathogenesis of various brain disorders and their symptoms. Equally difficult are the various potential routes of cure by drug and electrostimulation therapies. This is mainly because experimental studies are usually carried out to isolate the effects of a single mechanism and do not investigate the interactions of many mechanisms. This leads to a set of results that are conflicting, very difficult to interpret, or not integrated in a unified framework. Thus, in order to achieve patient-specific treatment and gain more knowledge about disorders, data-driven (Machine learning/Deep learning) and theory driven (Mathematical and computational models) approaches are increasingly needed in medicine as it can deal with diverse data, such as cellular, imaging, ‘omics’ or drug discovery. On one hand machine learning allows computers to perform tasks (for instance, predicting the disease trajectory for a patient) without having to explicitly provide rules. ML algorithms learn to perform a task by adjusting their parameters using training data. On the other hand, mathematical and computational models emerge as invaluable tools, because they uncover the biological mechanisms of the disease by providing coherent conceptual frameworks that integrate many different spatial and temporal scales and resolutions that allow for observing and experimenting with the system as a whole. Computational modellers then have precise control of experimental conditions needed for the replicability of experimental results. Because the process takes place in a computer, the investigator can perform multiple virtual experiments by preparing and manipulating the system in precisely repeatable ways and observe every aspect of the system without interference. However, the gap between experimental, computational/mathematical modellers and data scientists is still large. Bridging this gap is one of the central goals of the workshop. During this proposed workshop we will focus on successful cooperation of theoreticians and experimentalists with an aim to promote similar new collaborations. We are convinced that real progress can only be made via truly cross-disciplinary interactions. In our opinion, this is the best guarantee to create an atmosphere where ideas to stimulate interdisciplinary research in this field can blossom. We expect that the mutual interaction will contribute to addressing the right questions and to share the experimental and theoretical results obtained so far.

The workshop will invite contributions focusing on:

• • Data processing, image and signal reconstruction and enhancement and cross-modality synthesis
• • Extraction of biomarkers from ‘omics’, signal and imaging data
• • Machine learning algorithms in disease detection and diagnosis
• • Machine learning algorithms for prediction of disease evolution
• • Multi-scale, multi-level models of disease understanding
• • Computational/mathematical models of drug and stimulation treatments and therapies
• • Translation research on the dynamics of diseases into treatments for age related illnesses
• • Development of a culture of mathematical and computational modelling of brain diseases that will
• benefit those in clinical practice.

For these topics, emphasis will be given on the types of architectures and data to be used. Future trends will
be highlighted and guidelines to bridge the gap between research studies and the clinical route will be
provided.