Table 1 Ex vivo plaque characterisation with OCT compared with histology

Study	Plaque Type			Comments
Study	Lipid	Fibrous	Calcific	Comments
Yabushita et al (2002)
OCT	90/92	96/97	79/97	357 segments from 90 cadavers (162 aortas, 105 carotids, 90 coronaries). Training set of 50 used.
Kume et al (2006)
OCT	85/94	96/88	79/99	166 coronary sections from 40 cadavers. No training set. Few lipid plaques misclassified as fibrous or calcific.
IVUS	59*/97	88/86	98/96
Kawasaki et al (2006)
OCT	95/98	98/94	100/100	128 segments from 17 cadavers. No training set (OCT readers from Yabushita study performed OCT analysis).
IVUS	67/95	93/61	100/99
IB-IVUS	84/97	94/84	100/99
Rieber et al (2006)
OCT	77/94	67/97	64/88	17 segments from eight cadavers. Imaging with arteries in situ. Frequent sampling with 1 mm sections analysed in quadrants rather than as a whole.
IVUS	10*/96	63/59*	76/98
Manfrini et al (2006)
OCT	45/83†	83/82	68/76	79 coronary sections from 15 cadavers. 11 sections as training set. Difficulty of distinguishing between lipid and calcification and imaging deep lipid and calcium deposits were main issues.

Numbers are (%) sensitivity/specificity for detecting each plaque type using OCT, IVUS and IB-IVUS compared with histological controls. All studies except Manfrini et al used similar histological and OCT criteria. Manfrini et al used the AHA histological classification (†type IV and Va lesions being the equivalent of lipid plaque) with an adaptation of the previous OCT criteria. Reiber et al was the only study where imaging was performed on explanted hearts with arteries in situ (making coregistration with histology following explanation potentially difficult).
*Indicates a significant difference in comparison between OCT and IVUS (p<0.05).
IB-IVUS, integrated backscatter IVUS; IVUS, intravascular ultrasound; OCT, optical coherence tomography.