They apply an AI model to anticipate the risk of sepsis after surgery

Madrid. Researchers from the Bioengineering, Biomaterials and Nanomedicine area of ​​the Red Biomedical Research Center and the Infectious Diseases area at the University Clinical Hospital of Valladolid (Castilla y León) have applied an explainable artificial intelligence (XAI) model to identify and prioritize genetic information that increases the risk of developing sepsis after surgery.

Sepsis is a serious complication caused by an uncontrolled response of the body to an infection, usually of bacterial origin. It is the most severe form of sepsis and has a mortality rate that ranges between 10 and 20 percent, reaching 40 percent in cases of shock septic. On a global scale, sepsis causes around 11 million deaths each year, of which 17 thousand occur in Spain.

The work, published in Frontiershas analyzed data from a genome-wide association analysis that included genetic information from 753 patients who developed sepsis after surgery and 3,500 population controls. Using an explainable AI model, the researchers were able to not only predict the risk of sepsis, but also prioritize the genetic variants with the greatest contribution to this risk.

Thanks to this approach, the study identified genetic variations in the PRIM2, RBSN and SYNPR genes with functional, regulatory and clinical implications. In addition, genes related to key biological processes such as the regulation of gene expression, DNA replication, cell signaling, cell proliferation and cardiac dysfunction were detected.

As noted by the researchers, determining these genetic variants through preoperative blood tests could become a useful tool to improve risk stratification in surgical patients, facilitate early detection of postoperative sepsis and guide more personalized clinical interventions, with the aim of improving clinical outcomes and patient survival.

The study points out that more research will be necessary, including analysis in vitro e aliveas well as complementary studies in cohorts comprising patients with and without sepsis undergoing major surgery to optimally evaluate genetic factors contributing to predisposition to sepsis and provide external validation of exploratory findings.