Unfortunately, that is one of the biggest challenges. As once I heard: everyone likes #FAIRdata, but no one likes making their #data #FAIR

#HealthRI2020

#FAIR data is a key aspect to be efficient in learning from #data.

#HealthRI2020

Furthermore: what are the difficulties of learning from #data?

#HealthRI2020

Some difficulties that the #NHR needs to deal with. I can recognize a few myself when working with the #NKR

#HealthRI2020

A patient in an RCT in cardiology costs ~1,000-2,000 € per patient/year 😮

#HealthRI2020

Often, randomized control trials are considered the gold standard. However, they have some shortcomings that definitely worth pointing out

#HealthRI2020

Interesting plot (with data from the Swedish Heart Registry): transparency can improve care

#HealthRI2020

.@UMCUtrecht’s hub/“carrousel” to reduce bias and make models more generalizable

#HealthRI2020

*Predictions* for #AI: by 2060, computers will overrun humans

(I always take this numbers with a pinch of salt, though)

#HealthRI2020

Now for the next session: Better Data, Better Care

#HealthRI2020

.@IKNL’s @gijsg explaining the importance of the #pht @personalhealt11 for national and international collaborations

#HealthRI2020

Now for the unveiling of the #personalhealthtrain #pht @personalhealt11. If you have never heard of it, take a look at the clip here https://youtu.be/mktAtHmy-FM

#HealthRI2020

Something that I always forget: how is @Health__RI structured? Its CEO @LeoneFlikweert gives an overview

#HealthRI2020

.@Health__RI tries to tackle these issues on three pillars

1. Collective voice: align stakeholders
2. Infrastructure: #FAIR data and #federatedLearning
3. Providing services: portals, tools, etc.

#HealthRI2020

For example, #algorithms make it to the clinic very slowly. One of the reasons is that they are usually trained (and useful) locally only. They usually don’t generalize well

#HealthRI2020

#Data fuels science for society. However, #healthcare data poses its own set of challenges, which makes this process particularly complex

#HealthRI2020

.@Health__RI tries to bridge the gap between care and research. This will alsp make the transition from care to prevention

#HealthRI2020

It is not only about defragmenting data, but bringing people and ideas together. In the #Netherlands, university medical centers are nicely positioned to take the lead in this endeavour

#HealthRI2020

Hospitals, municipalities, individuals... never have we had so much data. However, it is a fragmented landscape, which we need to change

#HealthRI2020

The increase of life expectancy all around the world raises the need for a change of how we spend money and resources

#HealthRI2020

Now for a general update of @Health__RI and future perspectives

There’s a real need and push for a well-defined infrastructure for sharing data and research

#HealthRI2020

What do people think about human #genetics #research? Americans are actually quite positive

#HealthRI2020

Are we ready to make #genomics accessible to all medical professionals in all specialties? Not really. In the US, there are 5 medical genetecists/1 M inhabitants. There’s a big effort in training more and more professionals, though

#HealthRI2020

However, we need to also be critical. While databases have grown in volume, they fail on diversity representation

#HealthRI2020

What if your doctor offered genetic testing as a diagnosis tool? Hot question 🔥. First tests are being deployed in clinical environments, such as the US and UK

#HealthRI2020

Genomic studies have increases largely in size and scope. All these increases are largely due to us getting better at sharing data (centralized or #federated), storage, algorithms, pipelines, and #infrastructure

#HealthRI2020

Let’s take a look at #BioMe, NY’s biobank. Interesting place: >35% of people were born outside NY. They have >60,000 patients

#HealthRI2020

Number of biobanks have also increased. Here, 21 biobanks shown that have >20,000 patients

#HealthRI2020

This explosion also includes companies that offer #genetic profiling in the US and Europe

#HealthRI2020

Cost of Human Genome has rebased #MooresLaw. In other words, genome tech has grown (and cheapened) incredibly fast) #HealthRI2020.