Table 1 Summary of the literature reported reading time for CT scans with and without DL-CAD assistance. Reading time in mean (± SD) or median (IQR)
Study | Country or region | CT modality | DL-CAD product | Reading time (s) | Time difference (s) | P value | |
|---|---|---|---|---|---|---|---|
No DL-CAD | DL-CAD | ||||||
DL-CAD as a second reader | |||||||
Hsu (2021) [15] | Taiwan | Low dose CT | ClearReadCT (Riverain Technologies, Miamisburg, OH, USA) | 156 ± 34 | 197 ± 46 | 41 [95% CI:39 to 44] | < 0.001 |
Vassallo (2019) [16] | Italy | CT | M5L lung CAD on-demand, INFN | 296 ± 80 | 329 ± 83 | 33 | < 0.05 |
DL-CAD as a concurrent reader | |||||||
Hempel (2022) [17] | The Netherlands | CT | Veye Chest v2.15.3, Aidence B.V., Amsterdam, NL) | Reader 1: 226.4 ± 113.2 Reader 2: 320.8 ± 164.2 | Reader 1: 150.8 ± 74.2 Reader 2: 184.2 ± 125.3 | Reader 1: − 75.6 Reader 2: − 136.6 | < 0.001 |
Jacobs (2021) [18] | USA | Low dose CT | Veolity Lung CAD, version 1.5, MeVis Medical Solutions | 160 (96–245) | 86 (51–141) | − 64 (IQR: − 8 to − 137) | < 0.001 |
Hsu (2021) [15] | Taiwan | Low dose CT | ClearReadCT system (Riverain Technologies, Miamisburg, OH, USA) | 156 ± 34 | 124 ± 25 | − 32 [95% CI: − 30 to − 44] | < 0.001 |
Kozuka (2020) [19] | Japan | CT | InferRead CT Lung | 186 | 168 | − 18 | Not provided |
Lo (2018) [20] | USA | Low dose CT | ClearRead CT Vessel Suppression, Riverain Technologies | 127.2 | 95.6 | − 31.6 (95% CI: − 15.9 to − 47.4) | < 0.01 |
Pooled | – | – | – | 162 (95% CI: 111–212) | 118 (95% CI: 82–154) | 77 (95% CI: 47–107) | – |