AI outperforms conventional diagnosis for certain types of heart attacks

The AI-based method correctly identified obstructive MI in 84per cent of cases. Sensitivity was 77per cent, specificity was 99 per cent, and the negative predictive value was 98 per cent. There were 27 false negatives (2 per cent) and 17 false positives (1 per cent).

from Top Health News | Latest Healthcare Sector & Healthcare Industry news, Information and Updates: ET HealthWorld : ETHealthworld.com https://ift.tt/ArVDhgd
via