Specificity
Of the patients who do not have a condition, the proportion that the model correctly clears. High specificity means few false alarms.
A department choosing between two threshold settings for a chest X-ray tool that flags possible lung cancer is offered 95% specificity on one and 98% on the other. In a year that includes about 10,000 outpatient films from patients without cancer, that is roughly 500 versus 200 healthy patients flagged. Every flag the reporting radiologist cannot dismiss becomes a CT, with the radiation dose, cost and worry that follow. Ask what the 98% setting gives up in sensitivity, then weigh the 300 extra flags at 95% against the cancers that setting would additionally catch.
Terms like this come up in real clinical scenarios across the HelloAI courses: bite-sized modules with verifiable certificates. An account takes one minute, no password needed.
Sign in →