Sensitivity indices are quantitative measures used in sensitivity analysis to determine how variations in input parameters impact the output of a mathematical model. They help identify which inputs are most influential, guiding model simplification, uncertainty reduction, and decision-making processes.