Table 3.

Summary of results of the observer study and quantitative measurements. For the observer study, estimated marginal means for every quality parameter, reconstruction algorithm and significance versus DLR are given, with 95% confidence intervals in brackets. For the quantitative measurements, mean and standard deviation are shown

		Hybrid-IR	MBIR	DLR	DLR vs. Hybrid-IR	DLR vs. MBIR
Perceived ratings
	Noise	3.21 [3.12–3.30]	2.78 [2.69–2.87]	3.55 [3.46–3.64]	p < 0.001	p < 0.001
	Sharpness	3.55 [3.45–3.65]	2.84 [2.74–2.94]	3.62 [3.52–3.72]	p = 0.947	p < 0.001
	Natural appearance	3.63 [3.54–3.73]	2.87 [2.78–2.96]	3.54 [3.45–3.63]	p = 0.446	p < 0.001
	Grey-white matter differentiation	3.40 [3.31–3.50]	2.85 [2.75–2.95]	3.59 [3.49–3.69]	p = 0.027	p < 0.001
	Artefacts	3.45 [3.35–3.54]	2.86 [2.76–2.96]	3.40 [3.30–3.49]	p = 1.000	p < 0.001
	Overall perceived image quality	3.53 [3.44–3.63]	2.87 [2.28–2.96]	3.67 [3.57–3.76]	p = 0.154	p < 0.001
Quantitative analysis
	Standard deviation in lateral ventricle	6.4 ± 1.1	6.2 ± 0.7	5.6 ± 1.0	p < 0.001	p < 0.001
	Signal-difference-to-noise-ratio	1.9 ± 0.6	2.0 ± 0.6	2.4 ± 0.7	p < 0.001	p < 0.001
	Standard deviation in fourth ventricle*	6.3 ± 1.5	6.2 ± 0.7	5.3 ± 1.0	p < 0.001	p < 0.001
Reconstruction time
	Average reconstruction time [s]	27	176	44

*Two patients were excluded from these measurements as the fourth ventricle was filled with blood