A bacterial
antimicrobial resistance prediction method and
system, a hierarchical configuration unit divides resistance prediction elements into
patient risk levels, microbial feature levels, regional
prevalence levels, clinical rule levels and
drug property levels, each level includes initial attribute values and level-specific dynamic
rule sets; a
data processing unit acquires real-time clinical data, each level independently processes data through its level-specific dynamic rule set, and converts initial attribute values into target attribute values; a decision superposition unit superimposes the target attribute values of each level according to the weight priority of each level to generate a bacterial
antimicrobial resistance prediction result and an
antibacterial drug recommendation scheme; a report output unit outputs a visual report. The present application realizes precise, interpretable and highly adaptive bacterial
antimicrobial resistance prediction through
hierarchical design and rule monitoring separation architecture, while reducing the dependence on high-performance equipment and adapting to the use scenarios of primary medical institutions.