Automatic plaque characterization employing quantitative and multicontrast MRI

High-resolution and accelerated multi-parametric mapping with automated characterization of vessel disease using intravascular MRI

BACKGROUND

Atherosclerosis is prevalent in cardiovascular disease, but present imaging modalities have limited capabilities for characterizing lesion stage, progression and response to intervention. This study tests whether intravascular magnetic resonance imaging (IVMRI) measures of relaxation times (T1, T2) and proton density (PD) in a clinical 3 Tesla scanner could characterize vessel disease, and evaluates a practical strategy for accelerated quantification.

METHODS

IVMRI was performed in fresh human artery segments and swine vessels in vivo, using fast multi-parametric sequences, 1-2 mm diameter loopless antennae and 200-300 μm resolution. T1, T2 and PD data were used to train a machine learning classifier (support vector machine, SVM) to automatically classify normal vessel, and early or advanced disease, using histology for validation. Disease identification using the SVM was tested with receiver operating characteristic curves. To expedite acquisition of T1, T2 and PD data for vessel characterization, the linear algebraic method ('SLAM') was modified to accommodate the antenna's highly-nonuniform sensitivity, and used to provide average T1, T2 and PD measurements from compartments of normal and pathological tissue segmented from high-resolution images at acceleration factors of R ≤ 18-fold. The results were validated using compartment-average measures derived from the high-resolution scans.

RESULTS

The SVM accurately classified ~80% of samples into the three disease classes. The 'area-under-the-curve' was 0.96 for detecting disease in 248 samples, with T1 providing the best discrimination. SLAM T1, T2 and PD measures for R ≤ 10 were indistinguishable from the true means of segmented tissue compartments.

CONCLUSION

High-resolution IVMRI measures of T1, T2 and PD with a trained SVM can automatically classify normal, early and advanced atherosclerosis with high sensitivity and specificity. Replacing relaxometric MRI with SLAM yields good estimates of T1, T2 and PD an order-of-magnitude faster to facilitate IVMRI-based characterization of vessel disease.

Spatio-temporal texture (SpTeT) for distinguishing vulnerable from stable atherosclerotic plaque on dynamic contrast enhancement (DCE) MRI in a rabbit model

PURPOSE

To develop a new spatio-temporal texture (SpTeT) based method for distinguishing vulnerable versus stable atherosclerotic plaques on DCE-MRI using a rabbit model of atherothrombosis.

METHODS

Aortic atherosclerosis was induced in 20 New Zealand White rabbits by cholesterol diet and endothelial denudation. MRI was performed before (pretrigger) and after (posttrigger) inducing plaque disruption with Russell's-viper-venom and histamine. Of the 30 vascular targets (segments) under histology analysis, 16 contained thrombus (vulnerable) and 14 did not (stable). A total of 352 voxel-wise computerized SpTeT features, including 192 Gabor, 36 Kirsch, 12 Sobel, 52 Haralick, and 60 first-order textural features, were extracted on DCE-MRI to capture subtle texture changes in the plaques over the course of contrast uptake. Different combinations of SpTeT feature sets, in which the features were ranked by a minimum-redundancy-maximum-relevance feature selection technique, were evaluated via a random forest classifier. A 500 iterative 2-fold cross validation was performed for discriminating the vulnerable atherosclerotic plaque and stable atherosclerotic plaque on per voxel basis. Four quantitative metrics were utilized to measure the classification results in separating between vulnerable and stable plaques.

RESULTS

The quantitative results show that the combination of five classes of SpTeT features can distinguish between vulnerable (disrupted plaques with an overlying thrombus) and stable plaques with the best AUC values of 0.9631 ± 0.0088, accuracy of 89.98% ± 0.57%, sensitivity of 83.71% ± 1.71%, and specificity of 94.55% ± 0.48%.

CONCLUSIONS

Vulnerable and stable plaque can be distinguished by SpTeT based features. The SpTeT features, following validation on larger datasets, could be established as effective and reliable imaging biomarkers for noninvasively assessing atherosclerotic risk.

In-vivo quantitative T2 mapping of carotid arteries in atherosclerotic patients: segmentation and T2 measurement of plaque components

BACKGROUND

Atherosclerotic plaques in carotid arteries can be characterized in-vivo by multicontrast cardiovascular magnetic resonance (CMR), which has been thoroughly validated with histology. However, the non-quantitative nature of multicontrast CMR and the need for extensive post-acquisition interpretation limit the widespread clinical application of in-vivo CMR plaque characterization. Quantitative T2 mapping is a promising alternative since it can provide absolute physical measurements of plaque components that can be standardized among different CMR systems and widely adopted in multi-centre studies. The purpose of this study was to investigate the use of in-vivo T2 mapping for atherosclerotic plaque characterization by performing American Heart Association (AHA) plaque type classification, segmenting carotid T2 maps and measuring in-vivo T2 values of plaque components.

METHODS

The carotid arteries of 15 atherosclerotic patients (11 males, 71 ± 10 years) were imaged at 3 T using the conventional multicontrast protocol and Multiple-Spin-Echo (Multi-SE). T2 maps of carotid arteries were generated by mono-exponential fitting to the series of images acquired by Multi-SE using nonlinear least-squares regression. Two reviewers independently classified carotid plaque types following the CMR-modified AHA scheme, one using multicontrast CMR and the other using T2 maps and time-of-flight (TOF) angiography. A semi-automated method based on Bayes classifiers segmented the T2 maps of carotid arteries into 4 classes: calcification, lipid-rich necrotic core (LRNC), fibrous tissue and recent IPH. Mean ± SD of the T2 values of voxels classified as LRNC, fibrous tissue and recent IPH were calculated.

RESULTS

In 37 images of carotid arteries from 15 patients, AHA plaque type classified by multicontrast CMR and by T2 maps (+ TOF) showed good agreement (76% of matching classifications and Cohen's κ = 0.68). The T2 maps of 14 normal arteries were used to measure T2 of tunica intima and media (T2 = 54 ± 13 ms). From 11865 voxels in the T2 maps of 15 arteries with advanced atherosclerosis, 2394 voxels were classified by the segmentation algorithm as LRNC (T2 = 37 ± 5 ms) and 7511 voxels as fibrous tissue (T2 = 56 ± 9 ms); 192 voxels were identified as calcification and one recent IPH (236 voxels, T2 = 107 ± 25 ms) was detected on T2 maps and confirmed by multicontrast CMR.

CONCLUSIONS

This carotid CMR study shows the potential of in-vivo T2 mapping for atherosclerotic plaque characterization. Agreement between AHA plaque types classified by T2 maps (+TOF) and by conventional multicontrast CMR was good, and T2 measured in-vivo in LRNC, fibrous tissue and recent IPH demonstrated the ability to discriminate plaque components on T2 maps.

Ex vivo study of carotid endarterectomy specimens: quantitative relaxation times within atherosclerotic plaque tissues

PURPOSE

Previous studies reporting relaxation times within atherosclerotic plaque have typically used dedicated small-bore high-field systems and small sample sizes. This study reports quantitative T(1), T(2) and T(2) relaxation times within plaque tissue at 1.5 T using spatially co-matched histology to determine tissue constituents.

METHODS

Ten carotid endarterectomy specimens were removed from patients with advanced atherosclerosis. Imaging was performed on a 1.5-T whole-body scanner using a custom built 10-mm diameter receive-only solenoid coil. A protocol was defined to allow subsequent computation of T(1), T(2) and T(2) relaxation times using multi-flip angle spoiled gradient echo, multi-echo fast spin echo and multi-echo gradient echo sequences, respectively. The specimens were subsequently processed for histology and individually sectioned into 2-mm blocks to allow subsequent co-registration. Each imaging sequence was imported into in-house software and displayed alongside the digitized histology sections. Regions of interest were defined to demarcate fibrous cap, connective tissue and lipid/necrotic core at matched slice-locations. Relaxation times were calculated using Levenberg-Marquardt's least squares curve fitting algorithm. A linear-mixed effect model was applied to account for multiple measurements from the same patient and establish if there was a statistically significant difference between the plaque tissue constituents.

RESULTS

T(2) and T(2) relaxation times were statistically different between all plaque tissues (P=.026 and P=.002 respectively) [T(2): lipid/necrotic core was lower 47 ± 13.7 ms than connective tissue (67 ± 22.5 ms) and fibrous cap (60 ± 13.2 ms); T(2): fibrous cap was higher (48 ± 15.5 ms) than connective tissue (19 ± 10.6 ms) and lipid/necrotic core (24 ± 8.2 ms)]. T(1) relaxation times were not significantly different (P=.287) [T(1): Fibrous cap: 933 ± 271.9 ms; connective tissue (1002 ± 272.9 ms) and lipid/necrotic core (1044 ± 304.0 ms)]. We were unable to demarcate hemorrhage and calcium following histology processing.

CONCLUSIONS

This study demonstrates that there is a significant difference between qT(2) and qT(2) in plaque tissues types. Derivation of quantitative relaxation times shows promise for determining plaque tissue constituents.