Hacker News Reader

Comment by Gimpei | Hacker News Reader

With the caveat that I do not know the specifics of this case, I agree in principle with giving health companies access to large amounts of data like this. Research like this could save a huge amount of lives, but it needs lots of data, which is currently quiet difficult to collect. Ideally there'd be some sort of central, deidentified repository of data that all researchers can access for free. But in the absence of that, I'll take what I can get. I've lost too many people to cancer; I'd sacrifice a lot to see better diagnostics. Maybe I'm strange, but I don't care if Google or any other healthcare has access to all of the x-rays, mris, and blood tests that I've ever done provided that they're deidentified.

View on HN

The central repository is TCGA and dbGAP. It's pseudoanonymized but some datasets are restricted acces. When I worked for Google, we applied for access and downloaded the whole thing (petabytes) and then got an email from NCBI that we had to delete it all. Our idea was that we could serve a copy of it "for free" for researchers (internal to Google or external) to use in the cloud (without having to redownload it). This morphed a bit and became Public Datasets with User-Pays access. And AWS hosts it as well. Far more people use the data on AWS than anywhere else.

> some datasets are restricted acces.

That does not impress me. China and North Korea will get it.

Researchers don't need access to the data. They need to be able to run analyses on the data and get results. This can be achieved without handing over the data and in the UK the NHS actually built a system to do this. And then the corrupt Tory government handed the data over to Palantir anyway.

What exactly do you mean? Very few, if any, hosts of health research data have the data, network, and compute infrastracture to make use of their data. It should be in the cloud so it's actually useful.

I mean that the researchers submit jobs to the data owners who run the jobs, possibly in the cloud, and return the results to the researchers. The researchers don't need direct access to the data. I, along with millions of others, have opted out of every NHS digital initiative because I don't trust them. And my instincts were right because the data was given to Palantir. But if I was given the choice I would allow my data to be used in the privacy-respecting system described above.

I honestly don't think that model works. What if you want to join data between two datasets from two providers? Who pays for the cloud? What if the researchers want to do more featurization, making snapshot views for performance?

Having looked at similar contracts (specifically UofC's CrimeLab running analysis of police data against UofC hospital xray data), the pairing is required to be done on a host such that the xray data never leaves the UofC imaging infrastructure. Contractually, at least. Not exactly sure where your questions are coming from.

yeah, that happens. the researchers hate it (the service provided is often inferior to their own in-house setup) and it's mainly to retain some level of funding for the lab, rather than benefiting the most users. Also, university infrastructure is far less secure than a cloud account run by a cloud provider.

Reply on news.ycombinator.com