Table 4.

Phase contrast MRI accuracy in quantifying renal blood flow

Study	Methods	Reference technique	Accuracy results
Ku et al. [47]	In vitro	LDV, theoretical calculations	Excellent accuracy, good correlation (no reported values)
	Recirculating steady flow phantom (25.4 mm diameter)
Debatin et al. [50]	In vitro	True flow	Flow error: 42.1 ± 10.3% (16-frame cine PC-MRI), − 10.4 ± 17.3% (1-frame 2D PC-MRI), − 2.4 ± 2.5% (6-frame triads PC-MRI)
	Flow phantom
Siegel et al. [45]	In vitro	LDV	Velocity error < 30% (provided high SNR, low TE, thick slices)
	Stenotic flow phantom
Lee et al. [44]	In vitro	Doppler US	Peak systolic velocity: SEE = 5.3 cm/s (fast PC vs Doppler US), 4.6 cm/s (cine PC vs Doppler US); minimum diastolic velocity: r = 0.74 (fast PC vs Doppler US)
	Flow phantoms (10 mm diameter) with 0, 50, 75% stenosis
Hoppe et al. [43]	In vitro	Doppler US guidewire	Flow: r = 0.95
	Flow phantoms with varying concentric stenosis
King et al. [35]	In vitro	True flow	Steady flow: R2 = (0.77–0.99), relative flow error = (1.5–112)% according to pixel resolution [(0.8–3.6) mm]; pulsatile flow: error = (0.6–4.1)% (5 mm diameter) (1.4–13.8)% (2 mm diameter)
	Steady flow phantoms (3–11 mm diameter) + pulsatile flow phantoms (2 and 5 mm diameter)
Hollnagel et al. [46]	In vitro	LDV, CFD	Maximum velocity: RMSE = 6.55% (internal cerebral artery, PC-MRI vs LDV), 8.02% (internal cerebral artery, PC-MRI vs CFD), 9.34% (anterior cerebral artery, PC-MRI vs LDV), 10.35% (anterior cerebral artery, PC-MRI vs CFD)
	Flow phantom
Dambreville et al. [41]	In vitro	True flow	Steady flow: CCC = 0.992, mean flow error = 6.0 mL/min; pulsative flow: CCC = 0.996, mean flow error = − 6.7 mL/min
	Steady flow phantoms (3–11 mm diameter) + pulsatile-flow phantoms with additional 2.1 mm diameter channel
Khodarahmi et al. [48]	In vitro	PIV	Steady flow: CC > 0.99, pulsatile flow: CC > 0.96
	Flow phantom with varying concentric stenosis
Spithoven et al. [42]	In vitro	True flow	Flow: CC = 0.969 (6–8 mm diameter)
	Flow phantoms (5–8 mm diameter)
Sommer et al. [19]	In vivo	PAH clearance	RBF: r = 0.63 (artery), r = 0.76 (vein)
	9 HVs
Lundin et al. [53]	In vivo	PAH clearance	Good agreement (no reported values)
	14 HVs
Wolf et al. [52]	In vivo	PAH clearance	RBF error = 39 mL/min (95% CI − 100 to 177) (Venc = 100), 69 mL/min (95% CI − 31 to 169) (Venc = 150);
	10 HVs
Debatin et al. [50]	In vivo	PAH clearance	Mean RBF error = 2.8 ± 7.1% (6-frame triads PC-MRI), 28.5 ± 28.2% (16-frame), − 11.6 ± 14.9% (1-frame)
	8 HVs
Myers et al. [1]	In vivo	PAH clearance	RBF error = 20 mL/min (95% CI − 214 to 254), r = 0.91
	14 renal transplant recipients
Cortsen et al. [49]	In vivo	PAH clearance, 99mTc-DTPA scintigraphy	RBF: r = 0.68
	8 CKD patients
Sommer et al. [51]	In vivo	PAH clearance	Mean RBF error = (0–95) mL/min, RBF bias = (− 1.3 to 10)%, 95% CI =  ± (17.6–26.5)%, based on cine PC-MRI sequences (segmented k space, rapid spiral)
	18 HVs
de Haan et al. [55]	In vivo	133Xenon Washout	RBF: r = 0.69; CC = 0.51
	71 kidneys from patients with suspected renovascular hypertension
Spithoven et al. [42]	In vivo	Hip clearance	RBF: r = 0.81
	21 ADPKD patients

PC, phase contrast; Venc, velocity encoding; PIV, particle image velocimetry; Hip, ¹³¹I-hippuran; SEE, standard error of the estimate; US, ultrasound; SNR, signal-to-noise ratio; ICC, intraclass correlation coefficient; RMSE, root mean squared error; CFD, computational fluid dynamics; LDV, laser Doppler velocimetry; RBF, renal blood flow; HVs, healthy volunteers; ADPKD, autosomal dominant polycystic kidney disease; CKD, chronic kidney disease; PAH, para-aminohippurate; CV, coefficient of variation; CI, confidence interval; TE, echo time; CCC, concordance correlation coefficient