Skip to main content
. 2022 Nov;63(11):1768–1774. doi: 10.2967/jnumed.121.263686

FIGURE 2.

FIGURE 2.

Sensitivity and specificity of reader interpretation with and without deep learning (DL). Summed stress score > 3 was considered abnormal. Two thresholds were established for CAD-DL to match sensitivity of average reader specificity with (dark bars) and without (light bars) access to DL predictions. *P < 0.05 reader with DL compared with reader without DL, #P < 0.001 CAD-DL operating autonomously compared with all readers.