Optical coherence elastography based on inverse compositional Gauss-Newton digital volume correlation with second-order shape function.

A digital volume correlation (DVC)-based optical coherence elastography (OCE) method with inverse compositional Gauss-Newton (IC-GN) algorithm and second-order shape function is presented in this study. The systematic measurement errors of displacement and strain from our OCE method were less than 0.2 voxel and 4 × 10-4, respectively. Second-order shape function could better match complex deformation and decrease speckle rigidity-induced error. Compared to conventional methods, our OCE method could track a larger strain range up to 0.095 and reduce relative error by 30-50%. This OCE method has the potential to become an effective tool in characterising mechanical properties of biological tissue.

Plaque Longitudinal Heterogeneity in Morphology, Property, and Mechanobiology.

BACKGROUND AND PURPOSE: The hemodynamic environment of an atherosclerotic plaque varies along the longitudinal direction. Investigating the changes in plaque morphology and its biomechanical environment along the longitudinal direction and their correlations will enhance our understanding of plaque progression and arterial remodeling. METHODS: Six male patients with carotid stenosis >70% were recruited. Multisequence high-resolution MRI was performed at the carotid bifurcation. Carotid endarterectomy was performed following MRI, and the plaque tissue was collected for histological and mechanical testing. Patient-specific biomechanical modeling and simulations were conducted to calculate the mechanical stresses (wall shear stress [WSS] and von Mises stress [VMS]). Changes in plaque cross-sectional morphology, WSS, and VMS as well as their correlations were evaluated. RESULTS: Positive correlations were found between % stenosis and % inflammation (MA) (p = 0.019), % lipid area and % MA (p = 0.026), and % calcification area and VMS (p = 0.007). Negative correlations were found between VMS and % stenosis (p = 0.028) and VMS and average WSS (p = 0.034). Moreover, the peak stresses and neovessels were found to be in the shoulder regions. High-stress concentrations were found in the interface regions of the calcification and surrounding tissue, thereby increasing plaque vulnerability. CONCLUSIONS: Correlations between the morphology and stresses suggest that arterial remodeling is a dynamic interaction between mechanical environment and plaque progression resulting in plaque heterogeneity. Our finding indicates that plaque heterogeneity is associated with plaque progression and can be combined with mechanical stresses for identifying high-risk plaques.

Case Report: Evaluating Biomechanical Risk Factors in Carotid Stenosis by Patient-Specific Fluid-Structural Interaction Biomechanical Analysis.

BACKGROUND: Carotid atherosclerosis is one of the main underlying inducements of stroke, which is a leading cause of disability. The morphological feature and biomechanical environment have been found to play important roles in atherosclerotic plaque progression. However, the biomechanics in each patient's blood vessel is complicated and unique. METHOD: To analyse the biomechanical risk of the patient-specific carotid stenosis, this study used the fluid-structure interaction (FSI) computational biomechanical model. This model coupled both structural and hemodynamic analysis. Two patients with carotid stenosis planned for carotid endarterectomy were included in this study. The 3D models of carotid bifurcation were reconstructed using our in-house-developed protocol based on multisequence magnetic resonance imaging (MRI) data. Patient-specific flow and pressure waveforms were used in the computational analysis. Multiple biomechanical risk factors including structural and hemodynamic stresses were employed in post-processing to assess the plaque vulnerability. RESULTS: Significant difference in morphological and biomechanical conditions between 2 patients was observed. Patient I had a large lipid core and serve stenosis at carotid bulb. The stenosis changed the cross-sectional shape of the lumen. The blood flow pattern changed consequently and led to a complex biomechanical environment. The FSI results suggested a potential plaque progression may lead to a high-risk plaque, if no proper treatment was performed. The patient II had significant tandem stenosis at both common and internal carotid artery (CCA and ICA). From the results of biomechanical factors, both stenoses had a high potential of plaque progression. Especially for the plaque at ICA branch, the current 2 small plaques might further enlarge and merge as a large vulnerable plaque. The risk of plaque rupture would also increase. CONCLUSIONS: Computational biomechanical analysis is a useful tool to provide the biomechanical risk factors to help clinicians assess and predict the patient-specific plaque vulnerability. The FSI computational model coupling the structural and hemodynamic computational analysis, better replicates the in vivo biomechanical condition, which can provide multiple structural and flow-based risk factors to assess plaque vulnerability.

Low muscle mass and Charlson comorbidity index are risk factors for short-term postoperative prognosis of elderly patients with gastrointestinal tumor: a cross-sectional study.

BACKGROUND: Sarcopenia is one of the most frequent syndromes in older adults and one of its main characteristics is low muscle mass. Gastrointestinal tumor is a malignant disease with high incidence. This study aimed to investigate the risk factors of low muscle mass in older adults with gastrointestinal tumor, the prognostic indicators of and short-term outcomes after resection for gastrointestinal tumor, and to explore the relationship between low muscle mass and short-term postoperative prognosis. METHOD: A total of 247 older patients with gastrointestinal tumors who underwent radical resection in 2019 were included in this study. Relevant indexes were calculated using L3 slice image of computed tomography (CT) to evaluate low muscle mass. Short-term postoperative complications and length of stay were considered as short-term outcomes of this study. RESULTS: Advanced age, lower higher body mass index (BMI), lower hemoglobin, having history of abdominal surgery and higher visceral fat index (VFI) were risk factors of low muscle mass, while higher BMI and lower subcutaneous fat index (SFI) were protective factors of low muscle mass. Further multivariate logistic regression analysis showed that having history of abdominal surgery, advanced age and lower BMI were independent risk factors. Low muscle mass and higher Charlson comorbidity index were independent risk factors of short-term postoperative complications in older adults with gastrointestinal tumor. Higher Charlson comorbidity index gave rise to longer length of stay. CONCLUSIONS: Low muscle mass and higher Charlson comorbidity index predict poor short-term prognosis of older patients undergoing gastrointestinal tumor resection.

Fatigue characteristics of deformable mirrors caused by fatigue of lead zirconate titanate.

The correction ability of deformable mirrors (DMs) with different fatigued degrees has been studied by quantitatively analyzing the fitting wavefront, the times-diffraction-limit factor ß, and the wavefront power spectral density (PSD). The results indicate that the fatigue of DMs can be ignored when the DM continually works below 105 cycles, while the performance of the DM degrades gradually if the loading cycles of the DMs exceed 105. Consequently, the peak-valley (PV) value of the fitting wavefront of the DM decreases with the increase of the fatigue degree of the DM, resulting in the increase of the PV value of the corresponding residual wavefront. Meanwhile, the larger PV value of the distorted wavefront or the higher proportion of the high-spatial-frequency components gives rise to the larger residual wavefront after the existence of the fatigue of the PZT actuators, the worse corrected beam quality, and the lower correction ability of the DM. The correction ability of the DM composed of different piezoelectric ceramics has also been evaluated by the ß factor of the corrected laser beam, which indicates that for the given loading cycle, the fatigue degree and the correcting ability of the DM composed of different piezoelectric ceramics are different.

A survey on personal data cloud.

Personal data represent the e-history of a person and are of great significance to the person, but they are essentially produced and governed by various distributed services and there lacks a global and centralized view. In recent years, researchers pay attention to Personal Data Cloud (PDC) which aggregates the heterogeneous personal data scattered in different clouds into one cloud, so that a person could effectively store, acquire, and share their data. This paper makes a short survey on PDC research by summarizing related papers published in recent years. The concept, classification, and significance of personal data are elaborately introduced and then the semantics correlation and semantics representation of personal data are discussed. A multilayer reference architecture of PDC, including its core components and a real-world operational scenario showing how the reference architecture works, is introduced in detail. Existing commercial PDC products/prototypes are listed and compared from several perspectives. Five open issues to improve the shortcomings of current PDC research are put forward.

Quantifying irregular pulsation of intracranial aneurysms using 4D-CTA.

Recent studies have suggested that irregular pulsation of intracranial aneurysm during the cardiac cycle may be potentially associated with aneurysm rupture risk. However, there is a lack of quantification method for irregular pulsations. This study aims to quantify irregular pulsations by the displacement and strain distribution of the intracranial aneurysm surface during the cardiac cycle using four-dimensional CT angiographic image data. Four-dimensional CT angiography was performed in 8 patients. The image data of a cardiac cycle was divided into approximately 20 phases, and irregular pulsations were detected in four intracranial aneurysms by visual observation, and then the displacement and strain of the intracranial aneurysm was quantified using coherent point drift and finite element method. The displacement and strain were compared between aneurysms with irregular and normal pulsations in two different ways (total and stepwise). The stepwise first principal strain was significantly higher in aneurysms with irregular than normal pulsations (0.20±0.01 vs 0.16±0.02, p=0.033). It was found that the irregular pulsations in intracranial aneurysms usually occur during the consecutive ascending or descending phase of volume changes during the cardiac cycle. In addition, no statistically significant difference was found in the aneurysm volume changes over the cardiac cycle between the two groups. Our method can successfully quantify the displacement and strain changes in the intracranial aneurysm during the cardiac cycle, which may be proven to be a useful tool to quantify intracranial aneurysm deformability and aid in aneurysm rupture risk assessment.

Constrained estimation of intracranial aneurysm surface deformation using 4D-CTA.

BACKGROUND AND OBJECTIVE: Intracranial aneurysms are relatively common life-threatening diseases, and assessing aneurysm rupture risk and identifying the associated risk factors is essential. Parameters such as the Oscillatory Shear Index, Pressure Loss Coefficient, and Wall Shear Stress are reliable indicators of intracranial aneurysm development and rupture risk, but aneurysm surface irregular pulsation has also received attention in aneurysm rupture risk assessment. METHODS: The present paper proposed a new approach to estimate aneurysm surface deformation. This method transforms the estimation of aneurysm surface deformation into a constrained optimization problem, which minimizes the error between the displacement estimated by the model and the sparse data point displacements from the four-dimensional CT angiography (4D-CTA) imaging data. RESULTS: The effect of the number of sparse data points on the results has been discussed in both simulation and experimental results, and it shows that the proposed method can accurately estimate the surface deformation of intracranial aneurysms when using sufficient sparse data points. CONCLUSIONS: Due to a potential association between aneurysm rupture and surface irregular pulsation, the estimation of aneurysm surface deformation is needed. This paper proposed a method based on 4D-CTA imaging data, offering a novel solution for the estimation of intracranial aneurysm surface deformation.

MRI-based mechanical analysis of carotid atherosclerotic plaque using a material-property-mapping approach: A material-property-mapping method for plaque stress analysis.

BACKGROUND AND OBJECTIVE: Atherosclerosis is a major underlying cause of cardiovascular conditions. In order to understand the biomechanics involved in the generation and rupture of atherosclerotic plaques, numerical analysis methods have been widely used. However, several factors limit the practical use of this information in a clinical setting. One of the key challenges in finite element analysis (FEA) is the reconstruction of the structure and the generation of a mesh. The complexity of the shapes associated with carotid plaques, including multiple components, makes the generation of meshes for biomechanical computation a difficult and in some cases, an impossible task. To address these challenges, in this study, we propose a novel material-property-mapping method for carotid atherosclerotic plaque stress analysis that aims to simplify the process. METHODS: The different carotid plaque components were identified and segmented using magnetic resonance imaging (MRI). For the mapping method, this information was used in conjunction with an in-house code, which provided the coordinates for each pixel/voxel and tissue type within a predetermined region of interest. These coordinates were utilized to assign specific material properties to each element in the volume mesh which provides a region of transition. The proposed method was subsequently compared to the traditional method, which involves creating a composed mesh for the arterial wall and plaque components, based on its location and size. RESULTS: The comparison between the proposed material-property-mapping method and the traditional method was performed in 2D, 3D structural-only, and fluid-structure interaction (FSI) simulations in terms of stress, wall shear stress (WSS), time-averaged WSS (TAWSS), and oscillatory shear index (OSI). The stress contours from both methods were found to be similar, although the proposed method tended to produce lower local maximum stress values. The WSS contours were also in agreement between the two methods. The velocity contours generated by the proposed method were verified against phase-contrast magnetic resonance imaging (MRI) measurements, for a higher level of confidence. CONCLUSION: This study shows that a material-property-mapping method can effectively be used for analyzing the biomechanics of carotid plaques in a patient-specific manner. This approach has the potential to streamline the process of creating volume meshes for complex biological structures, such as carotid plaques, and to provide a more efficient and less labor-intensive method.

Differences in Carotid Artery Geometry and Flow Caused by Body Postural Changes and Physical Exercise.

Different body postures and physical exercises may lead to changes in arterial geometry and hemodynamics, which may be associated with the distribution of atherosclerosis lesions. This study was aimed at investigating potential geometric and hemodynamic changes of the carotid bifurcation in different body postures and after high-intensity interval training (HIIT) workouts. Three-dimensional vascular ultrasound (3DVUS) and Doppler ultrasound images were acquired for 21 healthy participants (aged 29 ± 6 y, 14 men and 7 women) in different body postures (sitting and three sleeping postures [supine, left lateral and right lateral]) and after physical exercises. The common carotid artery (CCA) and internal carotid artery (ICA) diameters of the left carotid artery were found to increase significantly from supine to left lateral (both p <0.05). CCA diameters (p < 0.05) and ICA/CCA diameter ratio (p < 0.01) of the left carotid artery changed significantly from supine to sitting. Significant differences in CCA peak systolic velocity (CCA PSV, p < 0.001), CCA end-diastolic velocity (CCA EDV, p < 0.001), CCA pulsatility index (CCA PI, p < 0.001) and maximum velocity-based wall shear stress at the CCA (WSS(max) at the CCA, p < 0.001) were identified in different postures. After physical exercises, significant increases were observed in the CCA diameter (p < 0.001), CCA PSV (p < 0.001), ICA PSV (p < 0.05), WSS(max) at the CCA (p < 0.001) and WSS(max) at the ICA (p < 0.05), as were significantly lower values of the CCA EDV (p < 0.01) and ICA/CCA PSV ratio (p < 0.05). Side-to-side differences were also detected in different postural change scenarios and after physical exercise; more significant differences were found to occur only in the left-sided carotid artery. Significant differences were identified under postural change and after physical exercise among healthy adults, suggesting that daily activity has an effect on the carotid bifurcation. These changes may be associated with formation and development of carotid atherosclerosis. Moreover, these side differences might be severe for patients and worth further attention in clinical practice.

Reproducibility of the computational fluid dynamic analysis of a cerebral aneurysm monitored over a decade.

Computational fluid dynamics (CFD) simulations are increasingly utilised to evaluate intracranial aneurysm (IA) haemodynamics to aid in the prediction of morphological changes and rupture risk. However, these models vary and differences in published results warrant the investigation of IA-CFD reproducibility. This study aims to explore sources of intra-team variability and determine its impact on the aneurysm morphology and CFD parameters. A team of four operators were given six sets of magnetic resonance angiography data spanning a decade from one patient with a middle cerebral aneurysm. All operators were given the same protocol and software for model reconstruction and numerical analysis. The morphology and haemodynamics of the operator models were then compared. The segmentation, smoothing factor, inlet and outflow branch lengths were found to cause intra-team variability. There was 80% reproducibility in the time-averaged wall shear stress distribution among operators with the major difference attributed to the level of smoothing. Based on these findings, it was concluded that the clinical applicability of CFD simulations may be feasible if a standardised segmentation protocol is developed. Moreover, when analysing the aneurysm shape change over a decade, it was noted that the co-existence of positive and negative values of the wall shear stress divergence (WSSD) contributed to the growth of a daughter sac.

Impact of cyclic bending on coronary hemodynamics.

It remains unknown that the degree of bias in computational fluid dynamics results without considering coronary cyclic bending. This study aims to investigate the influence of different rates of coronary cyclic bending on coronary hemodynamics. To model coronary bending, a multi-ring-controlled fluid-structural interaction model was designed. A coronary artery was simulated with various cyclic bending rates (0.5, 0.75 and 1 s, corresponding to heart rates of 120, 80 and 60 bpm) and compared against a stable model. The simulated results show that the hemodynamic parameters of vortex Q-criterion, temporal wall shear stress (WSS), time-averaged WSS (TaWSS) and oscillatory shear index (OSI) were sensitive to the changes in cyclic rate. A higher heart rate resulted in higher magnitude and larger variance in the hemodynamic parameters. Whereas, the values and distributions of flow velocity and relative residence time (RRT) did not show significant differences between different bending periods. This study suggests that a stable coronary model is not sufficient to represent the hemodynamics in a bending coronary artery. Different heart rate conditions were found to have significant impact on the hemodynamic parameters. Thus, cyclic bending should be considered to mimic the realistic hemodynamics in future patient-specific coronary hemodynamics studies.

The Need to Shift from Morphological to Structural Assessment for Carotid Plaque Vulnerability.

Degree of luminal stenosis is generally considered to be an important indicator for judging the risk of atherosclerosis burden. However, patients with the same or similar degree of stenosis may have significant differences in plaque morphology and biomechanical factors. This study investigated three patients with carotid atherosclerosis within a similar range of stenosis. Using our developed fluid-structure interaction (FSI) modelling method, this study analyzed and compared the morphological and biomechanical parameters of the three patients. Although their degrees of carotid stenosis were similar, the plaque components showed a significant difference. The distribution range of time-averaged wall shear stress (TAWSS) of patient 2 was wider than that of patient 1 and patient 3. Patient 2 also had a much smaller plaque stress compared to the other two patients. There were significant differences in TAWSS and plaque stresses among three patients. This study suggests that plaque vulnerability is not determined by a single morphological factor, but rather by the combined structure. It is necessary to transform the morphological assessment into a structural assessment of the risk of plaque rupture.

How getting twisted in scaffold design can promote bone regeneration: A fluid-structure interaction evaluation.

Bone tissue engineering (BTE) uses engineering principles to repair large bone defects, which requires effective mass transport ability of scaffolds to support cellular activities during bone regeneration. Since the implanted BTE scaffolds keep deforming under physiological loading which influences the fluid flow and mass transport within the scaffold and surrounding tissue, thus, scaffold design needs to consider the mass transport behavior under the physiological loading. This work proposed a novel twist scaffold, and its mass transport efficiency under physiological loading conditions was evaluated by a fluid-structure interaction analysis. The results showed that compared to the non-twist scaffold, the twist scaffold could form a rotating flow under the physiological loading, which enhanced the mass transport and generated more appropriate wall shear stress (WSS) to promote bone regeneration. This highlighted the better mass transport efficiency of the twist scaffold. Therefore, getting twist may be a promising design strategy for future BTE scaffolds, and the fluid-structure interaction approach may be a more reliable method for bone regeneration studies in either in vivo or in vitro systems.

The Hollow Porous Sphere Cell Carrier for the Dynamic Three-Dimensional Cell Culture.

Large-scale mammalian cell culture is essential in cell therapy, vaccine production, and the manufacturing of therapeutic protein drugs. Due to the adherent growth characteristic of most mammalian cell types, the combination of cell carrier and bioreactor is a common choice in large-scale mammalian cell culture. Current cell carriers developed by polymer crosslinking, lithography, or emulsion drops are unable to obtain a structure with uniformed porous structure and porous interior design, which results in an inhomogeneous culture condition for cells and therefore cannot ensure an optimal dynamic culture condition for cell proliferation, matrix production, and cell differentiation. In addition, the fluidic shear stress (a standard mechanical stimulation in bioreactor culture) and inner-carrier velocity (to ensure nutrient transport and waste exchange), which influence cell viability and growth, are not well-controlled/analyzed due to an irregular porous structure with these traditionally synthesized cell carriers. To solve these problems, we designed four types of hollow porous spheres (HPS, 1.0 cm diameter) with different porous structures. To investigate the impacts of porous structure on surface shear stress and inner velocity, computational fluid dynamics (CFD) simulations were conducted to analyze the liquid flow behavior in HPSs, based on which an optimal structure with minimal surface shear stress and best inner velocity was obtained and fabricated using fused deposition modeling three-dimensional (3D) printing technology. Inspired by the industrial large-scale culture system, a novel 3D dynamic culture system was then established using HPSs to seed the cells, which were then placed in a mini bioreactor on a tube roller. CFD analysis showed that under 0.1 m/s water flow, the shear stress at most surface areas from four HPSs was lower than 20 dynes/cm2, which suggests that the HPSs should provide protection against physical stress to the cells living on the scaffold surface. A dynamic cell seeding was developed and refined using the 3D culture system, which increased the 32% seeding efficiency of MC3T3 cells compared to the traditional static cell seeding method. The cell proliferation analysis demonstrated that HPSs could speed up cell growth in dynamic cell culture. The HPS with a honeycomb-like structure showed the highest inner pore velocity (CFD analysis) and achieved the fastest cell proliferation and the highest cell viability. Overall, our study, for the first time, developed a 3D printed HPS cell culture device with a uniformed porous structure, which can effectively facilitate cell adhesion and proliferation in the dynamic cultural environment, thereby could be considered an ideal carrier candidate.

Stroke risk evaluation for patients with atrial fibrillation: Insights from left atrial appendage with fluid-structure interaction analysis.

The majority of cardioembolic strokes in patients with non-valvular atrial fibrillation (NVAF) are resulted from clot formation in the left atrial appendage (LAA). Current stroke risk stratification is based on the overall risks estimated from demographic and clinical profiles but not on individual anatomy or physiology. We aim to explore the differences in LAA morphological and hemodynamic parameters by comparing patients with and without a stroke history. Thirty-nine patients with persistent NVAF were included. Of these, 17 patients without a stroke history (non-stroke group) were compared with 22 patients with a history of stroke (stroke group). Their LAA geometric models were first reconstructed, and the morphological parameters were then measured. Furthermore, their LAA hemodynamic parameters were calculated by fluid-structure interaction analysis. Moreover, particle residual rates (PRR) and blood renewal rates (BRR) analyses were also employed to characterize the thrombogenesis dynamics. The results showed that compared to the non-stroke group, the stroke group had significant smaller LAA tortuosity and LAA orifice area, and significantly lower LAA orifice velocities (0.16 ± 0.10 vs 0.15 ± 0.06 cm/s; p = 0.044), but higher PRR (14.58 ± 9.43 vs 9.25 ± 4.67; p = 0.040) and BRR (52.41 ± 18.11 vs 38.36 ± 24.07; p = 0.044). These LAA morphological and hemodynamic parameters may be used to assess stroke risk in patients with NVAF.

Impact of stent malapposition on intracoronary flow dynamics: An optical coherence tomography-based patient-specific study.

Percutaneous coronary intervention with stent implantation has emerged as a popular approach to treat coronary artery stenosis. Stent malapposition (SM), also referred as incomplete stent apposition, could reduce stent tissue coverage and hence increase the risk of late stent thrombosis. The objective of this study was to investigate the impact of SM on intracoronary flow dynamics by combining optical coherence tomography (OCT) image-based model reconstruction and computational analysis. Firstly, a stenosed coronary artery model was reconstructed from OCT and angiography imaging data of a patient. Two structural analyses were carried out to simulate two types of coronary artery stent implantations: a fully-apposed (FA) case and a SM case. Then, based on the two deformed coronary geometries, two computational fluid dynamics (CFD) analyses were performed to evaluate the differences of hemodynamic metrics between the FA and the SM cases, including wall shear stress (WSS), time-averaged WSS (TWSS), oscillatory shear index (OSI), WSS gradient (WSSG), time-averaged WSSG (TWSSG), and relative residence time (RRT). The results indicated that maximum flow velocity was higher in the SM case than that of the FA case, due to the incomplete expansion of the stent and artery. Moreover, the SM case had a lower percentage of areas of adverse WSS (< 0.5 Pa) and RRT (> 10/Pa) but a higher percentage of areas of adverse OSI (> 0.1) and WSSG (> 5000 Pa/m). Specifically, the differences of OSI, WSSG, and RRT between the two cases were relatively small. It was suggested that SM might not be responsible for negative hemodynamic metrics which would further result in stent thrombosis on the basis of the present specific model.

Evaluating the Impact of Calcification on Plaque Vulnerability from the Aspect of Mechanical Interaction Between Blood Flow and Artery Based on MRI.

Acute cerebral ischemic events and thrombosis are associated with the rupture/erosion of carotid atherosclerotic plaques. The aim of the present study was to determine the impact of calcification deposition on the wall shear stress (WSS) and stresses within the plaques using 3D fluid-structure interaction (FSI) models. Six patients with calcified carotid atherosclerosis underwent multisequence magnetic resonance imaging (MRI) and were divided into three groups according to the calcification volume. To evaluate the role of the calcification deposition on the stresses, the calcification content was replaced by lipids and arterial tissue, respectively. By comparing the results from the simulation with calcification, and when changing it to lipids there was a significant increment in the stresses at the fibrous cap (p = 0.004). Instead, by changing it to arterial tissue, there was no significant difference (p = 0.07). The calcification shapes that presented the highest stresses were thin concave arc-shaped (AS1) and thin convex arc-shaped (AS3), with mean stress values of 107 ± 54.2 and 99.6 ± 23.4 kPa, respectively. It was also observed that, the calcification shape has more influence on the level of stress than its distance to the lumen. Higher WSS values were associated with the presence of calcification. Calcification shape plays an important role in producing high stresses in the plaque. This work further clarifies the impact of calcification on plaque vulnerability.

Automated classification of coronary plaque calcification in OCT pullbacks with 3D deep neural networks.

SIGNIFICANCE: Detection and characterization of coronary atherosclerotic plaques often need reviews of a large number of optical coherence tomography (OCT) imaging slices to make a clinical decision. However, it is a challenge to manually review all the slices and consider the interrelationship between adjacent slices. APPROACH: Inspired by the recent success of deep convolutional network on the classification of medical images, we proposed a ResNet-3D network for classification of coronary plaque calcification in OCT pullbacks. The ResNet-3D network was initialized with a trained ResNet-50 network and a three-dimensional convolution filter filled with zeros padding and non-zeros padding with a convolutional filter. To retrain ResNet-50, we used a dataset of ∼4860 OCT images, derived by 18 entire pullbacks from different patients. In addition, we investigated a two-phase training method to address the data imbalance. For an improved performance, we evaluated different input sizes for the ResNet-3D network, such as 3, 5, and 7 OCT slices. Furthermore, we integrated all ResNet-3D results by majority voting. RESULTS: A comparative analysis proved the effectiveness of the proposed ResNet-3D networks against ResNet-2D network in the OCT dataset. The classification performance (F1-scores = 94 % for non-zeros padding and F1-score = 96 % for zeros padding) demonstrated the potential of convolutional neural networks (CNNs) in classifying plaque calcification. CONCLUSIONS: This work may provide a foundation for further work in extending the CNN to voxel segmentation, which may lead to a supportive diagnostic tool for assessment of coronary plaque vulnerability.

Atherosclerotic Plaque Tissue Characterization: An OCT-Based Machine Learning Algorithm With ex vivo Validation.

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.