You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
thanks for this tool. I read the Aphrodite documentation and found that it has 2 modes for Feature extraction which are denoted as 1 and 2 in settings.R file
1 - From first keyword appearance to last keyword appearance (entire EHR duration)
Q1) Let's say a person is diagnosed with AMI on Apr 1st, 2005 and his last record on our EHR dataset is Dec 10, 2015. So, we get all the features of his (lab, procedures, drugs etc) as frequency till his last record in EHR. This is clear and I understand
Q2) But for mode 2, where you look at 10 years of his past history (from date of 1st keyword appearance) to EXPERIMENTAL USER UNDER CARE
Here, I didn't understand what did you mean by EXPERIMENTAL USER UNDER CARE? Would you be able to help us understand this? does it mean we can configure like 2 years, 3 years from 1st keyword appearance?
The text was updated successfully, but these errors were encountered:
Hello,
thanks for this tool. I read the Aphrodite documentation and found that it has 2 modes for Feature extraction which are denoted as 1 and 2 in
settings.R
file1 - From first keyword appearance to last keyword appearance (entire EHR duration)
Q1) Let's say a person is diagnosed with AMI on Apr 1st, 2005 and his last record on our EHR dataset is Dec 10, 2015. So, we get all the features of his (lab, procedures, drugs etc) as frequency till his last record in EHR. This is clear and I understand
Q2) But for mode 2, where you look at 10 years of his past history (from date of 1st keyword appearance) to EXPERIMENTAL USER UNDER CARE
Here, I didn't understand what did you mean by EXPERIMENTAL USER UNDER CARE? Would you be able to help us understand this? does it mean we can configure like 2 years, 3 years from 1st keyword appearance?
The text was updated successfully, but these errors were encountered: