Table 6.
Accuracy of the severity prediction module of the DDAE.
| Configuration | Lows | Mids | Highs |
|---|---|---|---|
| Flat | 0.87 ± 0.13 | N/A | N/A |
| Sloping | 0.93 ± 0.02 | N/A | 0.96 ± 0.01 |
| Precipitous | 0.95 ± 0.03 | N/A | 0.97 ± 0.02 |
| Reverse sloping | 0.98 ± 0.05 | N/A | 0.97 ± 0.05 |
| Cookie bite | 0.98 ± 0.04 | 1.00 ± 0.00 | 0.97 ± 0.05 |
| Reverse cookie bite | 0.91 ± 0.09 | 0.99 ± 0.03 | 0.93 ± 0.06 |
| Notched | 0.98 ± 0.02 | 0.72 ± 0.02* | 0.54 ± 0.17 |
| Atypical | 0.81 ± 0.16 | 0.66 ± 0.30 | 0.71 ± 0.13 |
*This descriptor represents frequencies most susceptible to host audiometric notches (i.e. 3,000 Hz, 4,000 Hz and 6,000 Hz), and corresponds to the deepest threshold in the notch.