"-omics sciences" and "deep learning and the future of care" are now public (but in Italian only, for the time being)

 You might remember our announcement from a few months ago about our partners' contribution to the Master ARTE (artificial intelligence and telemedicine) of the University of Parma.

The lectures contributed then as part of the program have been held under embargo to give enrolled students a reasonable advance access to content... However they are now offered to everyone in keeping with our spirit. We are glad to share the videos here below for your evaluation:

1) -omics sciences

2) Deep Learning and the future of care

Lowering barriers to bottom-up innovation and collaboration is more important than optimizing perfomances/efficiency

I have recently read this twitter post by John Carlos Baez and immediately something else came to my mind: the piece "The counter-intuitive rise of Python in scientific computing" arguing that if one tool (python in the specific case) feels arguably technically inferior to alternatives, especially in the hands of power users, then we must still reckon with the fact that it has become the de facto most common/popular language in scientific computing and data sciences. How is this possible, despite remaining quite puzzling to technicians, is a secret hidden in plain sight: we are talking of a tool that enormously lowers the barriers to access, and makes it rather easy to focus on exploration, reuse, and composition of preexisting work, despite at the cost of some performances and elegance, thus attracting all those users whose mission is not to specifically challenge and develop tools, but to use them in applied fields with their own very specialised knowledge/expertise to be kept up to date and integrated.

How is this interesting to us in ISOMERISM? We have to learn to navigate the proper shades of interdisciplinary collaboration if we are to be successful in the digitalisation of medicine. Technical elegance and performance optimisation should be the focus of those elements of the value chain that can 1) be relegated to the infrastructures, hiding away the need for hardcore specialised skills from users AND 2) be relegated to those elements of the value chain whose uncertainty within the system is limited/nihil (e.g. execution of computation, I/O, data preservation, energy optimisation, ...)

All other layers (more visible to the human stakeholders, whether care givers or patients) should be progressively more and more nuanced in the negotiation between technology focus and users' freedom, embracing the necessity to accept sub-optimality in order to accomodate uncertainty and evolution... and failure of the system, with the ensuing social actions needed to correct the trajectory. Composability and reuse, with low entry barriers, become a priority in such a sociotechnical infostructure, as success is not definable as technical compliance, but only evaluated on the meaning of the results produced by the interactions emerging within and around the system.

The EU Commission has usually favoured highly technical agendas, however some initiatives (i.e. PerMedCoE ) appear to have captured the gist of the reflection I just shared, at least judging by their early ventures, bringing fresh air in the landscape of platforms development for medicine... we will follow them to learn more, hoping to have found new allies