Search

There is a need to develop a validated algorithm for plaque characterization which can help to facilitate the standardization of optical coherence tomography (OCT) image interpretation of plaque morphology, and improve the efficiency and accuracy in the application of OCT imaging for the quantitative assessment of plaque vulnerability. In this study, a machine learning algorithm was implemented for characterization of atherosclerotic plaque components by intravascular OCT using ex vivo carotid plaque tissue samples. A total of 31 patients underwent carotid endarterectomy and the ex vivo carotid plaques were imaged with OCT. Optical parameter, texture features and relative position of pixels were extracted within the region of interest and then used to quantify the tissue characterization of plaque components. The potential of individual and combined feature set to discriminate tissue components was quantified using sensitivity, specificity, accuracy. The results show there was a lower classification accuracy in the calcified tissue than the fibrous tissue and lipid tissue. The pixel-wise classification accuracy obtained by the developed method, to characterize the fibrous, calcified and lipid tissue by comparing with histology, were 80.0, 62.0, and 83.1, respectively. The developed algorithm was capable of characterizing plaque components with an excellent accuracy using the combined feature set.

Atherosclerotic plaque rupture is a catastrophic event that contributes to mortality and long-term disability. A better understanding of the plaque mechanical behavior is essential for the identification of vulnerable plaques pre-rupture. Plaque is subjected to a natural dynamic mechanical environment under hemodynamic loading. Therefore, it is important to understand the mechanical response of plaque tissue under cyclic loading conditions. Moreover, experimental data of such mechanical properties are fundamental for more clinically relevant biomechanical modeling and numerical simulations for risk stratification. This study aims to experimentally and numerically characterize the stress-relaxation and cyclic mechanical behavior of carotid plaque tissue. Instron microtester equipped with a custom-developed setup was used for the experiments. Carotid plaque samples excised at endarterectomy were subjected to uniaxial tensile, stress-relaxation, and cyclic loading protocols. Thirty percent of the underlying load level obtained from the uniaxial tensile test results was used to determine the change in mechanical properties of the tissue over time under a controlled testing environment (Control tests). The stress-relaxation test data was used to calibrate the hyperelastic (neo-Hookean, Ogden, Yeoh) and linear viscoelastic (Prony series) material parameters. The normalized relaxation force increased initially and slowly stabilized toward the end of relaxation phase, highlighting the viscoelastic behavior. During the cyclic tests, there was a decrease in the peak force as a function of the cycle number indicating mechanical distension due to repeated loading that varied with different frequencies. The material also accumulated residual deformation, which increased with the cycle number. This trend showed softening behavior of the samples. The results of this preliminary study provide an enhanced understanding of in vivo stress-relaxation and cyclic behavior of the human atherosclerotic plaque tissue.

A patient-specific carotid bifurcation with tandem stenosis found at both internal carotid artery (ICA) and common carotid artery (CCA) was studied. The in vivo pre-carotid endarterectomy (pre-CEA) multi-spectral magnetic resonance imaging (MRI) were performed and in vitro post-CEA carotid plaque tissue sample was collected. MR imaging data and tissue sample staining histology were used to recognize the plaque components. Further, the computational fluid dynamics (CFD) were performed on four MR-based reconstructed 3D carotid bifurcation models (the patient-specific geometry with tandem stenosis and three presumptive geometries by removing the stenosis part). The flow and shear stress behavior affected by the tandem stenosis was analyzed. From the results of MR segmentation and histology analysis, plaque lipid pool and calcification were found at both ICA and CCA. From the result of CFD simulation, the flow shear stress behavior suggested the tandem stenosis as a more “dangerous” situation than a single-stenosis artery. Besides, the CFD results deduced that the stenosis at the CCA location formed initially and led to the subsequent formation of stenosis at ICA. This study suggests that when planning CEA, CFD simulation on the presumptive models could help clinicians to estimate the blood flow behavior after surgery. Particular attention should be paid to the case of tandem stenosis, as the local hemodynamic environment is more complex and treatment of one stenosis may lead to a variation in the hemodynamic loading on the second plaque, which may result in either a higher risk of plaque rupture or restenosis.

Hemiparkinsonism duration in patients with Parkinson's disease (PD) is a key time window to study early pathology of PD. We aimed to comprehensively explore the alterations of deformation and structural network in PD patients with hemiparkinsonism, which could potentially disclose the early biomarker for PD. Thirty-one PD patients with hemiparkinsonism and 37 age- and gender- matched normal controls were included in the present study. First of all, we normalized the left hemisphere of structural images as the contralateral side to the affected limbs. Deformation-based morphometry (DBM) was conducted to evaluate the brain atrophy and/or enlargement. structural networks were constructed by thresholding gray matter volume correlation matrices of 116 regions and analyzed using graph theoretical approaches (e.g., small-worldness, global, and nodal measures). Significantly decreased deformation values were observed in the temporoparietal regions like bilateral middle temporal gyri, ipsilateral precuneus and contralateral Rolandic operculum extending to supramarginal and postcentral gyri. Lower deformation values in contralateral middle temporal gyrus were negatively correlated with higher motor impairment which was dominated by akinesia/rigidity. Moreover, nodal reorganization of structural network mainly located in frontal, temporal, subcortex and cerebellum was bilaterally explored in PD patients with hemiparkinsonism. Increased nodal properties could be commonly observed in frontal lobes. Disruption of subcortex including basal ganglia and amygdala was detected by nodal local efficiency and nodal clustering coefficient. Twelve hubs, mainly from paralimbic-limbic and heteromodal networks, were disrupted and, alternatively, 14 hubs, most of which were located in frontal lobes, were additionally detected in PD patients with hemiparkinsonism. In conclusion, during hemiparkinsonism period, mild brain atrophy in the temporoparietal regions and widespread reorganization of structural network, e.g., enhanced frontal function and disruption of basal ganglia nodes, occurred in both hemispheres. With our data, we can also argue that MTG contralateral to the affected limbs (expressing clinically verified brain atrophy) might be a potential living biomarker to monitor disease progression. Therefore, the combination of DBM and structural network analyses can provide a comprehensive and sensitive evaluation for potential pathogenesis of early PD patients with hemiparkinsonism.

It is well known that disruption of basal ganglia function generates the motor symptoms in PD, however, these are presented in a heterogeneous manner; patients can be divided into tremor-dominant and akinesia/rigidity-dominant subtypes. To date, it is unknown if these differences in the motor symptoms could be explained by differences on the functional connectivity of basal ganglia with specific brain regions. In this study, we aimed to explore the alterations of the network-based and global functional connectivity linking to basal ganglia between the PD-TD and PD-AR patients. One hundred and six PD patients and 52 normal controls were recruited. According to the subscales of UPDRS motor scale, PD patients were divided into the PD-TD (n = 57) and PD-AR (n = 49) subtypes. We performed independent component analysis to identify basal ganglia network (BGN) involving connected brain regions having coactivation with basal ganglia. Eigenvector centrality mapping were processed and the eigenvector centrality in the subcortical component of BGN including the bilateral caudate nuclei, putamen, thalami and pallidum were extracted to measure the global connectivity. Compared with controls, whole PD patients or PD subtypes showed decreases of functional connectivity within the subcortical component of BGN, e.g., thalamus, pallidum and putamen. Compared with controls, decreased functional connectivity of precuneus and amygdala with basal ganglia was observed in the PD-TD while that of occipital lobule and precuneus was observed in the PD-AR. Compared with the PD-TD, significantly decreased functional connectivity between occipital lobule and cerebellum posterior lobule and basal ganglia was observed in the PD-AR, and such connectivity had positive correlations with tremor and negative correlations with akinesia/rigidity. We also observed enhanced global connectivity in the caudate nucleus and thalamus in the PD subtypes compared with controls. In conclusion, PD patients independent of motor subtypes consistently express similar alterations of functional connectivity within the subcortical component of BGN including network-based connectivity and global connectivity. Functional connectivity of cerebellum posterior lobule and occipital lobule with basal ganglia play important roles in the modulation of parkinsonian motor symptoms.