Influence of the arterial input sampling location on the diagnostic accuracy of cardiovascular magnetic resonance stress myocardial perfusion quantification

Table 3 ROC analysis for detection of abnormal segments by stress MBF and MPR

AIF sampling location	AUC (95 % CI)	Sensitivity, %	Specificity, %	Optimal cut-off	p-value
Stress MBF
Left atrium	0.83 (0.79–0.87)	86	73	2.24	< 0.001
Basal left ventricle	0.81 (0.77–0.85)	91	63	2.63	< 0.001
Mid-left ventricle	0.71 (0.66–0.75)	79	52	2.75	< 0.001
Apical left ventricle	0.78 (0.73–0.82)	91	59	2.65	< 0.001
Aortic root	0.90 (0.87–0.93)	80	85	2.02	< 0.001
MPR
Left atrium	0.82 (0.78–0.87)	85	68	2.27	< 0.001
Basal left ventricle	0.81 (0.77–0.85)	86	67	2.25	< 0.001
Mid-left ventricle	0.69 (0.64–0.73)	77	53	2.42	< 0.001
Apical left ventricle	0.80 (0.76–0.84)	90	70	2.03	< 0.001
Aortic root	0.83 (0.79–0.87)	75	81	1.88	< 0.001

Results based on 300 normal and 100 abnormal segments as diagnosed visually. The optimal cut-off indicates the value below which the segment is classified as positive (inclusive). The cut-off for stress MBF is measured in units of mL/g/min
AUC area under the curve, AIF arterial input function, CI confidence interval, MBF myocardial blood flow, MPR myocardial perfusion reserve

ISSN: 1532-429X