Table 1.
Description of the Datasets Used in the Deep Learning Models
| DIGS/ADAGES | MRCH | Iinan | Hiroshima | ACRIMA | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Normal | Glaucoma | Normal | Glaucoma | Normal | Glaucoma | Normal | Glaucoma | Normal | Glaucoma | |
| Training dataa | 1381/2601/8706 | 678/1293/5357 | 1768 | 1364 | — | — | — | — | — | — |
| Testing dataa | 180/319/483 | 90/150/276 | 49 | 61 | 110 | 95 | 75 | 91 | 309 | 396 |
| High myopia (%) | 0 | 2.9 | 27.5 | 35.6 | 0.9 | 3.2 | 0 | 0 | — | — |
| VF MD (dB)(95% CI) | −0.80(−1.02 to −0.58) | −3.32(−3.96 to −2.69) | — | −10.47(−12.60 to −8.34) | — | −4.31(−5.42 to −3.19) | −1.28(−2.04 to −0.53) | −13.54(−15.38 to −11.71) | — | — |
| Age (years)(95% CI) | 54.1(52.7 to 55.5) | 63.1(61.7 to 64.4) | 55.2(50.1 to 60.4) | 65.6(63.1 to 68.1) | 76.4(74.8 to 78.1) | 78.7(77.1 to 80.4) | 51.7(46.2 to 57.1) | 66.3(63.6 to 69.0) | — | — |
| Sex (% female) | 61.9 | 60.1 | 56.9 | 59.3 | 56.4 | 68.4 | 50.70 | 45.10 | — | — |
| Race (%) | ||||||||||
| African descent | 45.5 | 31.2 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Japanese descent | 1.2 | 5.4 | 100 | 100 | 100 | 100 | 100 | 100 | 0 | 0 |
| European descent | 51.6 | 62.7 | 0 | 0 | 0 | 0 | 0 | 0 | 100 | 100 |
| Other/unreported | 1.7 | 0.7 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
—, a measurement was unavailable for a given dataset.
a
The UCSD training and testing data are presented as the number of patients/eyes/scans.