Background Aortic Atheroma (AoA) is an independent risk factor for new and recurrent stroke. ulceration (≥ 2 mm) and BI 2536 mobility of AoA were assessed in the proximal (ascending and proximal arch) and distal (distal arch and descending) segments of thoracic aorta by a cardiologist to interpret the TEE and a radiologist to interpret the cMRI. There was good correlation between cMRI and TEE in measurement of plaque thickness in the proximal segments (R=0.73 p<0.0001) and the distal segments (R=0.81 p<0.0001) of the aortic arch (AA). cMRI had a high degree of accuracy in detecting measurable AoA (≥ Thbs2 1 mm) in the proximal segments (sensitivity 90% specificity 100%) as well as the distal segments (sensitivity 67% specificity 100%). cMRI also had a high degree of accuracy in detecting significant AoA (≥ 4 mm) in proximal segments (sensitivity 71% specificity 93%) as well as distal segments (sensitivity 71% specificity 100%). Conclusion The study showed a high degree of accuracy and correlation of AoA detected and measured by cMRI as compared to TEE in patients with recent stroke/TIA. This technique has limitations in detection of AoA ulceration and protocols assessing AoA mobility need to be developed. and and p-value <0.001) of the aortic arch (Figures 1 and ?and2).2). cMRI detected of proximal segment measurable AA disease ≥ 1 mm BI 2536 in thickness with 90% sensitivity and 100% specificity. In distal segments of the aortic arch the sensitivity was 67% and specificity was 100%. cMRI detected severe AoA disease (≥ 4 mm in thickness) in proximal segments of the aortic arch with 71% sensitivity and BI 2536 93% specificity. In the distal aortic arch the technique was 71% specific and 100% sensitive. cMRI was able to identify 89% of proximal aorta ulcerations and 64% of distal aorta ulcerations (Figure 3). The mobility of plaques could not be assessed with this protocol. Figure 1 AAPT correlation in proximal segments. Figure 2 AAPT correlation in distal segments. Figure 3 AA ulceration seen on cMRI and TEE. Discussion It has been shown that aortic atheromatous (AoA) disease may be found in 16-20% of patients with transient ischemic attack (TIA) or stroke and a number of prospective studies have confirmed a strong association between stroke and AoA plaques [7 8 17 Sen et al. [21] found a high rate of progression of aortic atheroma in 28% of ischemic stroke/TIA patients using transesophageal echocardiography (TEE) which is considered the gold standard for detection of AoA plaque in this setting [22]. TEE is widely available produces images of greater resolution than transthoracic echocardiography (TTE) and has high inter-reader reproducibility [23 24 However TEE has significant limitations. It is semi-invasive requires conscious sedation and carries risk of damage to the esophagus albeit low [24]. Patient tolerability may be a problem [25]; a relatively large number of individuals refused a follow-up TEE after a 12 month interval in one series [21]. Because of the semi-invasive nature of the study patients for TEE should be hemodynamically stable. Many patients with stroke have difficulty in swallowing [26] and BI 2536 thus may have difficulty swallowing the transducer. Air artifact and insonation angles can interfere with visualization in proximal [15 24 and distal aortic arch [27]. A failure to reliably assess the aortic arch because of insufficient visualization was noted in one study in about half the patients tested; it was shown that TEE may underestimate AoA disease [27]. Difficulties with plaque thickness measurement and gradation scoring by TEE at precise locations in baseline and follow-up measurements was a limiting factor in a study of progression of AoA disease and risk of stroke [28]. Cardiovascular magnetic resonance imaging (cMRI) of the thoracic aorta has been shown to have unique characteristics which may improve the potential for identifying BI 2536 the source of cryptogenic stroke; about 21% of a group of patients with stroke of undetermined etiology were found to have AoA disease with embolic potential which was not detected by TEE [27]. The development of new protocols for cMRI to.