I don't claim to understand this paper very well either, but based on a skim it seems the general idea is to do longitudinal blood testing on an individual leve

I don't claim to understand this paper very well either, but based on a skim it seems the general idea is to do longitudinal blood testing on an individual level to establish a baseline for blood parameters, to then be able to detect changes that can't easily be explained by normal physiological processes(e.g a sudden rise in hemoglobin). This is similar to what has been suggested before, but due to many confounding factors the individualised baseline may be necessary to avoid false positives.