Share your details and we'll email you our Publication
Authors:
Tanel Laisaar, Eran Choman, Alon Lanyado, Eitan Israeli, Andre Koit, Kaja Triin Laisaar.
Background:
LungFlag is a machine learning (ML) tool for calculating risk of pre-symptomatic lung
cancer, that uses electronic medical record (EMR) data as input. For this study,
LungFlag was used retrospectively on data collected from individuals who were referred
to screening LDCT.
Summary of Next Steps:
Use of LungFlag in Estonian lung cancer screening setup could be considered in the
following conditions:
1. To test whether LungFlag could outperform the PLCOm2012 (noRace) model when
screening all ever smokers.
2. Given availability of high-quality data, lung cancer risk assessment can be automated
for individuals with pre-existing sufficient EMR data avoiding the need for repeated
personal outreach.
3. Personalized risk assessment could motivate very high-risk individuals to attend initial
screening and repeated assessment to come back for yearly screening.
