Phoenix Rising supports the Millions Missing global day of protest
Phoenix Rising is delighted to support the demands being made in the ME/CFS community’s first-ever global day of protest …
Discuss the article on the Forums.

Aging.ai - deep neural network that can predict age from blood markers

Discussion in 'Other Health News and Research' started by undiagnosed, May 22, 2016.

  1. undiagnosed

    undiagnosed Senior Member

    Messages:
    220
    Likes:
    331
    United States
    I came across a post on reddit about a deep neural network that can predict age from blood markers. I did the basic test at the website and it predicted I was a 38 year old male. I am actually a 29 year old male. I am curious what results people here will get. I wonder if it was able to pick up on health issues and predicted a higher age or if it was just due to the prediction accuracy. I don't think I had any errors in the unit conversions. Anyways, I'd be interested to see other results.
     
    Theodore likes this.
  2. undiagnosed

    undiagnosed Senior Member

    Messages:
    220
    Likes:
    331
    United States
    FYI, in order to fill out the basic test, I needed results from the following panels:
    • Comprehensive Metabolic Panel (CMP)
    • Complete Blood Count (CBC)
    • Lipid Panel
    • Serum Protein Electrophoresis (SPE)
     
  3. undiagnosed

    undiagnosed Senior Member

    Messages:
    220
    Likes:
    331
    United States
    Also, if anyone needs help with unit conversions or anything, let me know.
     

See more popular forum discussions.

Share This Page