Towards better health & care: potential of Artificial Intelligence

It was the 24th of May and in Amsterdam - at Vrje University - the event Towards better health&care: the potential of AI has had place held by AMDS (you can find the official community group here.

But before dive into the topic - let’s have step back.

What AMDS is

AMDS stands for Amsterdam Medical Data Science and is a group (the official community group is on meetup.com) with over 2000 enthusiast doctors (and data scientists) that somehow work into the healthcare: meetings are said where doctors meet data scientists and data scientists meet doctors in order to talk about data and everything is around the medicine of tomorrow.

These two categories is are to speak the same language and is even more rare they may have the possibility to meet each other. Most of the time they speak a different language and a group cannot understand the other - and vice versa.

The group is kept on by Dr. Paul Elbers (Amsterdam UMC - Intensivist doctor) and by Dr. Mark Hoogendoorn (VU University Amsterdam - Mathematic) - with the financial support of Amsterdam UMC and of VU University Amsterdam.

The aim of the group is to empower the healthcare offer - tailing a medicine on the patient measure and to put the patient with her or him unique features at the center the journey - instead of at the conclusion of standardized processes.

I dati in medicina sono una grande ricchezza; la quantità con cui vengono giornalmente generati però non va di pari passo alla qualità con cui vengono raccolti e ciò si riflette su uno scarso uso che invece ne viene fatto.

Nei meetup, solitamente a cadenza mensile, si discute di medical data science - ovvero di progetti in corso, sviluppi, iniziative, opportunità che hanno a che fare con l’ambiente medico e dei dati.

Il tutto accompagnato da pizza.

The event

The event has been held in english - but it is not been only a meet-up - but a full-day event with workshops, postering on developing projects and speech about futures and theoretical developments.

But the thing outmost enlighted me was not the topics but speakers.

For the first time I head *doctors, colleagues speaking with the same languages usually informatics do - with an universal technical language. Hearing specific buzzwords such as random forest, logistic regression, support-vector machine spread along the speech with high-medicine-content topics moved me. Really.

The program

You can find the program has been sticked here.

Extremely interesting the keynote held by Sandjai Bhulai entitled Artificial Intelligence in healthcare … it matters! where he showed how a well fit AI algorithm based on the most common event-position that require an ambulance to be solved may reduce drastically the on-site time to arrive - overall in case of high requested ambulace period in specific day - just making them move along specific path to cover the territory has been left uncovered by another ambulance moved to an event-site.

But the most he showed how the the number of movements required to get a reasonable time-to-arriva onsite was quite feasable preserving the life-job quality for the with a human squad.

But as long they couldn’t implement the AI algo into the operations center they thought the operator to think as the AI may do with great results indeed.

Top!

Another interesting speech has been Applying AI to address the complex health care problem of falls in older persons held by Noman Dormosch & Bob van de Loo in which just allowing the patient to choose the hospital they wanted to be admited into the first aid phase ranked by the time they may had to wait according to the supposed health-problem (in the speech example: femorous fracture) has enough to make the choice of less ranked hospital acceptable by the patient where otherwise they may have complained about.

Into the speech Translating AI into Radiological practice by Henk Marquering (Radiologist) he tells us about the various winters the AI faced during his job-experience, the study of radiological images. The AI probably never will replace the Radiologist operator; but we can be sure the tech will improve, empower and get better the job. Also because in case of missed diagnosis - will never the AI algorithm be called in cause (into Tribunals - I mean - in Italy) and condemned?

Conclusions

I warmly suggest to attend such events for those who may be interested (Italian collegaues: give me a shout! 😁 - and let’s go togheter the next event!): they empower the self-awereness that the data analysys and its derivates may be a winner-player into the medical competences.

I’d like to close the post with a phrase always remarks this concept in a extra-layer: “I profile versus T profile”.