AI chatbots reading X-rays can be dangerously confident even when they’re wrong
The test ran 200 cases across 16 models and compared them against a panel of radiologists. Human experts scored 988.7 out of a possible 2,000 points. The best AI model hit 758. Human radiologists outperformed every tested AI model on the primary metric, which combines accuracy with the confidence level of each answer. | Image:…