A machine-learning algorithm can be used to predict patients' risk of contracting a potentially deadly hospital infection, according to a study covered by Digit.
The study was published in Nature Communications.
The research team, which specializes in microbiology and genetics at Aalto University and University of Helsinki in Finland, developed a machine-learning algorithm by combining large-scale population genomics and measurements of relevant features of life-threatening bacteria that come from hospital bugs.
The bacteria in the study, Staphylococcus epidermidis, which has been a part of normal human skin flora, has recently become a key source of infections through indwelling medical devices and surgeries like hip replacements. By using machine learning, the researchers successfully predicted patients' risk of developing an infection from the genomic features of the bacteria.
The research is still in its beginning stages and could not determine whether all the members of the S. epidermidis population that colonize human skin asymptomatically can induce high-risk infections. However, the findings open a path for future technology to be included in infection diagnosis by identifying high-risk genotypes of bacteria during a patient's surgery.