Table 3.
Overall image quality and image noise of clinical image
| Reconstruction method | ||||
|---|---|---|---|---|
| New DLR (AiCE Body sharp) | Conv DLR (AiCE Body) | Hybrid-IR | MBIR | |
| Overall image quality | 4.2 ± 0.4 (4.0–5.0) | 3.3 ± 0.4a (3.0–4.0) | 2.7 ± 0.4a (2.0–3.0) | 2.1 ± 0.2a (2.0–3.0) |
| Difference from New DLR | - | 1.0 ± 0.2 (0.0–1.0) | 1.5 ± 0.5 (1.0–2.0) | 2.2 ± 0.3 (2.0–3.0) |
| Image noiseb | 20.3 ± 1.8 | 17.8 ± 1.5 | 25.0 ± 2.7 | 23.7 ± 3.1 |
DLR, deep learning-based reconstruction; AiCE, advanced intelligent clear-IQ engine; Hybrid-IR, hybrid-iterative reconstruction; MBIR, model-based iterative reconstruction
Unless otherwise indicated, data are mean and standard deviation with range in parentheses
a
indicates a significant difference between the new DLR and each reconstruction algorithm (p < 0.0001)
b
Data are shown with mean and standard deviation