Intracranial artery atherosclerotic plaque is a major cause of stroke. The development of black blood MRI techniques has allowed the visualization of the intracranial artery wall in vivo. Composition of extracranial carotid plaque such as intra-plaque haemorrhage (IPH), fibrous cap and lipid core has been extensively studied as predictors of ischemic stroke¹. Characterizing intracranial plaque composition in vivo, however, is still challenging given its small size and the lack of histological validation. Previous post-mortem studies have reported the T1/T2/T2* relaxation times of plaque components at ultra-high field strengths (7T² and 17.6T³). However, most prior in vivo studies of intracranial plaque were undertaken using 3T clinical scanners, but relaxation times at 3T have not been reported. This study aims: 1) to quantify the relaxation times of intracranial plaque components ex vivo at 3T; and 2) to evaluate the ability of multi-contrast MRI for intracranial plaque type classification in comparison with histology.Study population: Specimens of the circle of Willis (CoW) were obtained from 20 cadavers (11 male; mean age 73.8 ±10.9 years) with atherosclerotic plaques. All specimens were rinsed with saline to remove blood clots. MRI acquisition: MRI scanning was undertaken in a Siemens 3T Skyra scanner using a loop coil (4cm diameter) within 24 hours after excising specimen. Specimens were embedded within Fomblin (a fluorinated fluid with no MR signal) during imaging. Quantitative T1/T2/T2* mapping sequences and multi-contrast fast-spin echo (FSE) sequences were acquired (Table 1). Image analysis: Specimens were stained with haematoxylin and eosin (H&E) and Masson trichrome. MRI images were co-registered with histology slices. Plaque components including IPH, fibrous cap, lipid core, fibrous tissue, calcification and healthy wall were segmented on histology, and their relaxation times were recorded on MR images. Plaque types were classified on histology by two experienced pathologists according to the American Heart Association (AHA) definitions 4. Two radiologists also classified the plaque types using multi-contrast MRI 4, blinded to the histology results. Statistics: Considering that multiple measurements were obtained from each subject, a linear mixed-effect model was used to assess the difference between the relaxation times of different components. Cohen’s kappa was used to estimate the agreement of plaque type classification using MRI vs. histology.In total, 53 atherosclerotic arteries were excised. 229 slices with matched MRI and histology locations were included in the analysis. Sample images are shown in Figures 1&2. Quantitative mapping results are shown in Table 2, and the agreement between MRI and histology is shown in Table 3. There are significant differences among the relaxation times of different components (p<0.05). T2 and T2* value of lipid core are lower than fibrous cap (p<0.001), but are comparable with fibrous tissue and healthy wall (p>0.05). Lipid core has less proton density compared with fibrous cap and fibrous tissue (p<0.001). Calcium has the lowest relaxation times and proton density. There is good agreement between MRI and histology for plaque type classification (κ = 0.69) with an overall accuracy of 80.7%. The sensitivity and specificity using MRI to identify fibro-lipid atheroma (type V-VI) is 95.8% and 77.1% respectively. Good inter-observer agreement was also found between two radiologists (κ = 0.77).To our knowledge, this is the first study to report the relaxation times of intracranial plaque components at 3T. The reported relaxation times can be used in protocol design and for the characterization of plaque components on multi-contrast imaging at 3T. We found lipid core has significantly lower T2 values than fibrous cap, such that they could be clearly distinguished on T2 weighted images. These findings agreed with previous carotid studies at 3T. However, previous intracranial plaque studies ^{2 3} did not show a significant difference between the T2 values of fibrous cap and lipid core at ultra-high field strengths. We also demonstrate that multi-contrast MRI can identify high-risk plaques with high sensitivity and specificity ex-vivo. Intracranial plaque components have distinct and different relaxation times at 3T. High-resolution MRI is able to characterize intracranial plaque composition and classify plaque types ex vivo at 3T. This study provides a basis for the development of in vivo MRI techniques to evaluate intracranial plaque vulnerability and improve risk stratification of patients.