Hi everyone!
EHF Director in AI and NLP Eray Ozkural:
What about EHF's approach to AI and NLP?
https://youtu.be/gVKMmyzDZ8M-----
How about the eHealth First approach to AI? We’re looking at a few possible applications, so I’m going to try to explain what we intend to accomplish. First is medical meta-analysis, looking through a good number of medical articles, and solving some medical problems using this data. This will have a conversational interface for ordinary users, and one that is intended for medical professionals where they can specify more detailed queries. This will help people to make an automated diagnosis based on their medical history and lab work and other sensor data, so we try to learn these things from medical articles, and we find the relevant publications and recommended course of action, basically, doing a lot of what a medical doctor does, of course in a more automated manner.
At first, we will make use of existing research, and integrate them into our architecture in a practical manner. And then, we will be adding a lot of things like deep learning, and we will automate the language analysis using deep learning approach, and this will make it a lot more adaptable, and a lot more powerful actually.
Deep Learning has started to take over a lot of things in natural language processing, so we’re trying to follow this technology, and we think this will work really well here. We can actually make a conversational engine like Alexa or Siri but one that’s a medical expert that knows all of those articles at a sufficient level. And we’re going to use the medical standards and conventions for understanding the text, so we can ensure sufficient understanding to answer some challenging queries.
You can ask about how to treat a general condition, or you can input your personalized data, lab work, sensor readings, medical images and it will infer what your condition is, it will diagnose you, and recommend a medical intervention much like what a doctor does. So, that’s the main goal.
Medical image analysis using deep learning will also be achieved by the same system. And, in a later stage, we will also integrate genomics data through deep learning, so it will be able to look at your genome sequence, proteomics, and other *omics data, so these are more advanced stuff, and more expensive stuff, but this is where AI can be really powerful, because personalizing treatment based on your genome is a more complex but potentially much more powerful problem. And then, it will also be able to look at medical images, use deep learning on medical images, to detect conditions like tumors or lesions.
So, these are the three basic capabilities of the EHF AI Platform. We think we can do a lot more, for instance, I’m considering a longevity application where we will recommend a longevity combination therapy, personalized, for a specific patient, because every patient will have different biomarkers to improve, and these personalized recommendations can thus be very useful.
These are some of the things we are going to do, and we are very excited about them!
-----
All the details are on the website
http://www.ehfirst.io/ and in the White Paper
http://www.ehfirst.io/#wp.
The WHITE LIST is launched!
Please use the link to fill in the form:
https://ehfirst.io/whitelistIt takes 5 minutes only.
NB! After submission the form all investors participate in the lottery, in which 20 randomly selected winners will receive an additional 2 000 EHF tokens.
In May 2018, EHF Pre-ICO will be launched.
Up to 1 500 000 EHF tokens for sale.
Price at launch 0.006 Ethers.
Daily price increment +0.25%.
Bonuses of up to 40% is available.
Take a look on
https://ehfirst.io