A novel multiparametric imaging approach to acute myocarditis using T2-mapping and CMR feature tracking

Table 4 Diagnostic performance of different CMR parameters for diagnosing suspected acute myocarditis

Parameter	AUC^a	Sensitivity [%]	Specificity [%]	PPV^b [%]	NPV^c [%]	Accuracy [%]
Lake Louise criteria
T2-Ratio (≥1.9)	0.58	40	77	87	25	48
Visual edema	0.72	43	100	100	31	55
EGEr^d	0.74	48	100	100	33	58
LGE^e	0.76	52	100	100	35	62
Strain parameters
LV^f longitudinal strain	0.68	73	59	88	36	70
LV longitudinal SR^g	0.66	60	71	89	31	62
LV circumferential strain	0.70	52	88	95	32	60
T2 parameters
Mean T2	0.75	69	82	94	40	71
MaxT2	0.79	87	82	95	61	86
Mean SD	0.70	72	71	91	39	71
MadSD	0.82	84	77	93	54	82
Multiparametric models without gadolinium
madSD + LV longitudinal strain	0.84	93	71	93	71	88
madSD + LV circumferential strain	0.84	93	82	95	74	91
madSD ≥1.8 ms + LV circumferential strain ≥ −25%	0.84	90	77	95	67	88
Multiparametric models with gadolinium
LLC^h	0.79	58	100	100	38	67
madSD + LV longitudinal strain + LGE	0.94	99	77	94	93	94
madSD + LV circumferential strain + LGE	0.93	97	82	96	88	94
madSD ≥1.8 ms + LV circumferential strain ≥ −25% + LGE pos.	0.94	97	77	96	88	94

^a AUC area under the curve, ^b PPV positive predictive value, ^c NPV negative predictive value, ^d EGEr Early Gadolinium Enhancement ratio, ^e LGE Late Gadolinium Enhancement, ^f LV left ventricle, ^gSR strain rate, ^h LLC Lake Louise criteria

ISSN: 1532-429X