Table 2.
Correct answers for each group of evaluators, in total and in detail
| Group | SD | SD + DD |
|---|---|---|
| % (95% CI) | % (95% CI) | |
| Dermatologists | 70.6 (65.5, 75.2) | 80.3 (75.7, 84.2) |
| Residents | 58.3 (49.4, 66.8) | 68.3 (59.6, 76.0) |
| Board-certified dermatologists | 70.8 (62.2, 78.2) | 79.2 (71.1, 85.5) |
| Experts in nail diseases | 85.0 (76.7, 90.7) | 96.0 (90.2, 98.4) |
| AI | 25.0 (16.8, 35.5) | 35.0 (25.5, 45.9) |
| ChatGPT | 25.0 (11.2, 46.9) | 30.0 (14.5, 51.9) |
| Grok | 15.0 (5.2, 36.0) | 25.0 (11.2, 46.9) |
| Gemini | 35.0 (18.1, 56.7) | 50.0 (29.9, 70.1) |
| Claude | 25.0 (11.2, 46.9) | 35.0 (18.1, 56.7) |
The bold values represent the aggregated results of the two main evaluator groups (all dermatologists combined and all AI models combined).
AI artificial intelligence algorithms, CI confidence interval, DD differential diagnosis, SD suspected diagnosis.