Histogram showing propensity scores as calculated from the processed data, coloured by whether there was a diagnosis of epilepsy (blue) or not (red).

Histogram showing propensity scores as calculated from the processed data, coloured by whether there was a diagnosis of epilepsy (blue) or not (red).

<p>There is a complete overlap–the range of scores for the two diagnostic classes overlaps completely.</p> <p>(TIFF)</p>

Saved in:

Bibliographic Details
Main Author:	Ioana Duta (18462981) (author)
Other Authors:	Symon M. Kariuki (10575863) (author), Anthony K. Ngugi (7906343) (author), Angelina Kakooza Mwesige (20714882) (author), Honorati Masanja (102691) (author), Daniel M. Mwanga (20037196) (author), Seth Owusu-Agyei (80063) (author), Ryan Wagner (257632) (author), J Helen Cross (10646762) (author), Josemir W. Sander (10946352) (author), Charles R. Newton (7243856) (author), Arjune Sen (73985) (author), Gabriel Davis Jones (8957774) (author)
Published:	2025
Subjects:	Biotechnology Cancer Hematology na &# 239 machine learning algorithms includes demographic information demonstrated substantial promise saharan african sites iteratively added sites demographic sites dataset clinical variables related active convulsive epilepsy trained using data xlink "> single evaluated model performance xlink "> trained personnel clinical tools sourced data local data predictive model model weights model performances validation analysis settings beyond preliminary screening novel approach limited resources intended deployment increasing number income countries generally worse external performance epilepsy live diagnostic process cultural context broader implementation aid screening 05 ).
Tags:	Add Tag No Tags, Be the first to tag this record!

Be the first to leave a comment!

Cannot write session to /tmp/vufind_sessions/sess_kpv5p17t7nhrglahu3918s8n51