Abstract
AbstractEffective response to vaccination requires activation of the innate immune system, triggering the synthesis of inflammatory cytokines. The subjective symptoms related to this, referred to as reactogenicity, affect a variable percentage of vaccinated people to different degrees, with evidence supporting a relationship between the severity of symptoms a person experiences and their eventual immune response. Wearable sensors allow for the identification of objective evidence of physiologic changes a person experiences in response to vaccine-induced inflammation, but as these changes are subtle, they can only be detected when an individual’s pre-vaccination normal variability is considered. We used a wearable torso sensor patch and a machine learning method of similarity-based modeling (SBM), which learns the dynamic interplay between multivariate input sources, to create a physiologic digital twin for 88 people receiving 104 vaccine doses. By effectively removing expected variations and leaving only vaccine-induced differences, we developed a multivariate digital biomarker that incorporates changes in multiple continuously monitored physiologic data streams to measure the degree and duration of vaccine induced inflammation. This objective measure correlated with subjective symptoms, and in a 20-person subset, both humoral and cellular immunogenicity.
Publisher
Cold Spring Harbor Laboratory