If you’re a physician who is seeing potential COVID-19 patients, please consider contributing to this wonderful open-source AI project by uploading de-identified x-rays to figure1.
You don’t have to identify yourself and you can even open the link from an anonymous browser.
If you have an x-ray of a patient who has had confirmed COVID-19 infection, you can take a screenshot on your computer screen, then crop to remove the patient name, and then upload it to the link above.
If you have access to such xrays, please help them. If not, forward that link to colleagues who can help. We will all benefit from the results of this project.
Open-source means that everyone can contribute to the project and it’s a not-for-profit endeavor.
More importantly, open-source means that others can access the data and build off of it. That’s one of my favorite things about data sharing in healthcare.
All of the data we upload to such platforms will then be available for researchers in the future who come along with more sophisticated tools or better ideas.
I’ve talked about Figure1 a lot in the past. It’s a place where clinical providers share their cases from all over the world. It’s very easy to use and you can learn a lot from other clinicians.
Sadly, there aren’t a lot of PA’s or NP’s on there. You guys need to get on there and represent.
You often will get responses fast. So it’s a great place to share a case and get an answer to a complicated situation.
AI in Medicine
The way AI is currently utilized it’s not really AI, but that doesn’t mean that it’s not helpful. AI refers to artificial intelligence, which is something we are experimenting with in healthcare but don’t have anything solid from it yet.
There are various depths to AI, from pattern recognition and algorithms – which is what we mostly have now – to deep learning using convoluted neural networks.
For any physician who is interested in getting into this space, this is definitely the time. I have a great course to help you get started in healthcare consulting, whether in AI or anything else in medicine.
I am still involved in my own healthcare consulting project with a team of machine learning experts. We are currently looking at this COVID-19 business and trying to see where we can add value.
Also, AI isn’t anything complicated, if you have a great idea for a project then pair up with a machine learning engineer, find or create your data set, and come up with a solution to a worthwhile problem in medicine.
The last thing I’ll mention is what data sets are. The data set is the complication of all the xrays you guys upload to that platform.
Those xrays are fed into a machine learning tool which will derive a set of conclusions from all the commonalities in those films.
This AI project will always have access to this data set in the future or make it available to others. With it, they can have the model evaluate a new, unread xray which might look normal to you or I but can be flagged as a possible COVID-19 case based on its subtle findings.
Google Cloud is offering free access to its sizable online Coronavirus database. No need to recruit your own patients in a study, all the data is collected an accessible on Google Cloud.