ML-driven segmentation of microvascular features during histological examination of tissue-engineered vascular grafts.

Introduction: The development of next-generation tissue-engineered medical devices such as tissue-engineered vascular grafts (TEVGs) is a leading trend in translational medicine. Microscopic examination is an indispensable part of animal experimentation, and histopathological analysis of regenerated tissue is crucial for assessing the outcomes of implanted medical devices. However, the objective quantification of regenerated tissues can be challenging due to their unusual and complex architecture. To address these challenges, research and development of advanced ML-driven tools for performing adequate histological analysis appears to be an extremely promising direction. Methods: We compiled a dataset of 104 representative whole slide images (WSIs) of TEVGs which were collected after a 6-month implantation into the sheep carotid artery. The histological examination aimed to analyze the patterns of vascular tissue regeneration in TEVGs in situ. Having performed an automated slicing of these WSIs by the Entropy Masker algorithm, we filtered and then manually annotated 1,401 patches to identify 9 histological features: arteriole lumen, arteriole media, arteriole adventitia, venule lumen, venule wall, capillary lumen, capillary wall, immune cells, and nerve trunks. To segment and quantify these features, we rigorously tuned and evaluated the performance of six deep learning models (U-Net, LinkNet, FPN, PSPNet, DeepLabV3, and MA-Net). Results: After rigorous hyperparameter optimization, all six deep learning models achieved mean Dice Similarity Coefficients (DSC) exceeding 0.823. Notably, FPN and PSPNet exhibited the fastest convergence rates. MA-Net stood out with the highest mean DSC of 0.875, demonstrating superior performance in arteriole segmentation. DeepLabV3 performed well in segmenting venous and capillary structures, while FPN exhibited proficiency in identifying immune cells and nerve trunks. An ensemble of these three models attained an average DSC of 0.889, surpassing their individual performances. Conclusion: This study showcases the potential of ML-driven segmentation in the analysis of histological images of tissue-engineered vascular grafts. Through the creation of a unique dataset and the optimization of deep neural network hyperparameters, we developed and validated an ensemble model, establishing an effective tool for detecting key histological features essential for understanding vascular tissue regeneration. These advances herald a significant improvement in ML-assisted workflows for tissue engineering research and development.

Enhancing decellularized vascular scaffolds with PVDF and PCL reinforcement: a fused deposition modeling approach.

Background: Decellularized xenogenic scaffolds represent a promising substrate for tissue-engineered vascular prostheses, particularly those with smaller diameters (<6 mm). Despite their benefits, a notable limitation presents itself during decellularization, namely, the diminished mechanical strength that introduces the risk of aneurysmal dilations in the early post-implantation period. This study introduces a strategy for modification the mechanical properties of these biological scaffolds through the forming of an external polymeric reinforcement via thermal extrusion. Methods: The study utilized scaffolds fabricated from bovine internal mammary arteries through decellularization and preservation. The scaffolds were divided into subgroups and reinforced with polymeric helices made of Polyvinylidene fluoride (PVDF) and Polycaprolactone (PCL), n = 5 for each. An experimental setup for external reinforcement coating was designed. Computed microtomography was employed to obtain accurate 3D models of the scaffolds. Mechanical properties were evaluated through in vitro uniaxial tension tests (Z50, Zwick/Roell, Germany), compliance evaluation and numerical simulations (Abaqus/CAE, Dassault Systemes, France) to investigate the effect of external reinforcement on aneurysm growth. Results: Using a double-layer helix for the reinforcement significantly enhanced the radial tensile strength of the scaffolds, increasing it up to 2.26 times. Yet, the comparison of vessel's compliance between two reinforced and the Control scaffolds within the physiological pressures range did not reveal any significant differences. Numerical simulation of aneurysm growth showed that thin-walled regions of the Control scaffold developed aneurysmal-type protrusions, bulging up to 0.7 mm, with a substantial degradation of mechanical properties. In contrast, both PVDF and PCL reinforced scaffolds did not exhibit significant property degradation, with deformations ranging 0.1-0.13 mm depending on the model, and a maximum decrease in the modulus of elasticity of 23%. Conclusion: The results of the study demonstrated that the external polymer helical reinforcement of decellularized scaffolds via thermal extrusion enables a controlled modification of mechanical properties, notably enhancing radial strength while maintaining sufficient compliance within the physiological pressure range. A series of in vitro tests demonstrated the consistency and potential of this approach for decellularized xenogenic scaffolds, a concept that had not been explored before.

Perfect prosthetic heart valve: generative design with machine learning, modeling, and optimization.

Majority of modern techniques for creating and optimizing the geometry of medical devices are based on a combination of computer-aided designs and the utility of the finite element method This approach, however, is limited by the number of geometries that can be investigated and by the time required for design optimization. To address this issue, we propose a generative design approach that combines machine learning (ML) methods and optimization algorithms. We evaluate eight different machine learning methods, including decision tree-based and boosting algorithms, neural networks, and ensembles. For optimal design, we investigate six state-of-the-art optimization algorithms, including Random Search, Tree-structured Parzen Estimator, CMA-ES-based algorithm, Nondominated Sorting Genetic Algorithm, Multiobjective Tree-structured Parzen Estimator, and Quasi-Monte Carlo Algorithm. In our study, we apply the proposed approach to study the generative design of a prosthetic heart valve (PHV). The design constraints of the prosthetic heart valve, including spatial requirements, materials, and manufacturing methods, are used as inputs, and the proposed approach produces a final design and a corresponding score to determine if the design is effective. Extensive testing leads to the conclusion that utilizing a combination of ensemble methods in conjunction with a Tree-structured Parzen Estimator or a Nondominated Sorting Genetic Algorithm is the most effective method in generating new designs with a relatively low error rate. Specifically, the Mean Absolute Percentage Error was found to be 11.8% and 10.2% for lumen and peak stress prediction respectively. Furthermore, it was observed that both optimization techniques result in design scores of approximately 95%. From both a scientific and applied perspective, this approach aims to select the most efficient geometry with given input parameters, which can then be prototyped and used for subsequent in vitro experiments. By proposing this approach, we believe it will replace or complement CAD-FEM-based modeling, thereby accelerating the design process and finding better designs within given constraints. The repository, which contains the essential components of the study, including curated source code, dataset, and trained models, is publicly available at https://github.com/ViacheslavDanilov/generative_design.

Embedding and Backscattered Scanning Electron Microscopy (EM-BSEM) Is Preferential over Immunophenotyping in Relation to Bioprosthetic Heart Valves.

Hitherto, calcified aortic valves (AVs) and failing bioprosthetic heart valves (BHVs) have been investigated by similar approaches, mostly limited to various immunostaining techniques. Having employed multiple immunostaining combinations, we demonstrated that AVs retain a well-defined cellular hierarchy even at severe stenosis, whilst BHVs were notable for the stochastic degradation of the extracellular matrix (ECM) and aggressive infiltration by ECM-digesting macrophages. Leukocytes (CD45+) comprised ≤10% cells in the AVs but were the predominant cell lineage in BHVs (≥80% cells). Albeit cells with uncertain immunophenotype were rarely encountered in the AVs (≤5% cells), they were commonly found in BHVs (≥80% cells). Whilst cell conversions in the AVs were limited to the endothelial-to-mesenchymal transition (represented by CD31+α-SMA+ cells) and the formation of endothelial-like (CD31+CD68+) cells at the AV surface, BHVs harboured numerous macrophages with a transitional phenotype, mostly CD45+CD31+, CD45+α-SMA+, and CD68+α-SMA+. In contrast to immunostaining, which was unable to predict cell function in the BHVs, our whole-specimen, nondestructive electron microscopy approach (EM-BSEM) was able to distinguish between quiescent and matrix-degrading macrophages, foam cells, and multinucleated giant cells to conduct the ultrastructural analysis of organelles and the ECM, and to preserve tissue integrity. Hence, we suggest EM-BSEM as a technique of choice for studying the cellular landscape of BHVs.

Cardiovascular and Renal Comorbidities Included into Neural Networks Predict the Outcome in COVID-19 Patients Admitted to an Intensive Care Unit: Three-Center, Cross-Validation, Age- and Sex-Matched Study.

Here, we performed a multicenter, age- and sex-matched study to compare the efficiency of various machine learning algorithms in the prediction of COVID-19 fatal outcomes and to develop sensitive, specific, and robust artificial intelligence tools for the prompt triage of patients with severe COVID-19 in the intensive care unit setting. In a challenge against other established machine learning algorithms (decision trees, random forests, extra trees, neural networks, k-nearest neighbors, and gradient boosting: XGBoost, LightGBM, and CatBoost) and multivariate logistic regression as a reference, neural networks demonstrated the highest sensitivity, sufficient specificity, and excellent robustness. Further, neural networks based on coronary artery disease/chronic heart failure, stage 3-5 chronic kidney disease, blood urea nitrogen, and C-reactive protein as the predictors exceeded 90% sensitivity and 80% specificity, reaching AUROC of 0.866 at primary cross-validation and 0.849 at secondary cross-validation on virtual samples generated by the bootstrapping procedure. These results underscore the impact of cardiovascular and renal comorbidities in the context of thrombotic complications characteristic of severe COVID-19. As aforementioned predictors can be obtained from the case histories or are inexpensive to be measured at admission to the intensive care unit, we suggest this predictor composition is useful for the triage of critically ill COVID-19 patients.

Polymeric Heart Valves Will Displace Mechanical and Tissue Heart Valves: A New Era for the Medical Devices.

The development of a novel artificial heart valve with outstanding durability and safety has remained a challenge since the first mechanical heart valve entered the market 65 years ago. Recent progress in high-molecular compounds opened new horizons in overcoming major drawbacks of mechanical and tissue heart valves (dysfunction and failure, tissue degradation, calcification, high immunogenic potential, and high risk of thrombosis), providing new insights into the development of an ideal artificial heart valve. Polymeric heart valves can best mimic the tissue-level mechanical behavior of the native valves. This review summarizes the evolution of polymeric heart valves and the state-of-the-art approaches to their development, fabrication, and manufacturing. The review discusses the biocompatibility and durability testing of previously investigated polymeric materials and presents the most recent developments, including the first human clinical trials of LifePolymer. New promising functional polymers, nanocomposite biomaterials, and valve designs are discussed in terms of their potential application in the development of an ideal polymeric heart valve. The superiority and inferiority of nanocomposite and hybrid materials to non-modified polymers are reported. The review proposes several concepts potentially suitable to address the above-mentioned challenges arising in the R&D of polymeric heart valves from the properties, structure, and surface of polymeric materials. Additive manufacturing, nanotechnology, anisotropy control, machine learning, and advanced modeling tools have given the green light to set new directions for polymeric heart valves.

Proteolytic Degradation Is a Major Contributor to Bioprosthetic Heart Valve Failure.

Background Whereas the risk factors for structural valve degeneration (SVD) of glutaraldehyde-treated bioprosthetic heart valves (BHVs) are well studied, those responsible for the failure of BHVs fixed with alternative next-generation chemicals remain largely unknown. This study aimed to investigate the reasons behind the development of SVD in ethylene glycol diglycidyl ether-treated BHVs. Methods and Results Ten ethylene glycol diglycidyl ether-treated BHVs excised because of SVD, and 5 calcified aortic valves (AVs) replaced with BHVs because of calcific AV disease were collected and their proteomic profile was deciphered. Then, BHVs and AVs were interrogated for immune cell infiltration, microbial contamination, distribution of matrix-degrading enzymes and their tissue inhibitors, lipid deposition, and calcification. In contrast with dysfunctional AVs, failing BHVs suffered from complement-driven neutrophil invasion, excessive proteolysis, unwanted coagulation, and lipid deposition. Neutrophil infiltration was triggered by an asymptomatic bacterial colonization of the prosthetic tissue. Neutrophil elastase, myeloblastin/proteinase 3, cathepsin G, and matrix metalloproteinases (MMPs; neutrophil-derived MMP-8 and plasma-derived MMP-9), were significantly overexpressed, while tissue inhibitors of metalloproteinases 1/2 were downregulated in the BHVs as compared with AVs, together indicative of unbalanced proteolysis in the failing BHVs. As opposed to other proteases, MMP-9 was mostly expressed in the disorganized prosthetic extracellular matrix, suggesting plasma-derived proteases as the primary culprit of SVD in ethylene glycol diglycidyl ether-treated BHVs. Hence, hemodynamic stress and progressive accumulation of proteases led to the extracellular matrix degeneration and dystrophic calcification, ultimately resulting in SVD. Conclusions Neutrophil- and plasma-derived proteases are responsible for the loss of BHV mechanical competence and need to be thwarted to prevent SVD.

Early Postoperative Immunothrombosis of Bioprosthetic Mitral Valve and Left Atrium: A Case Report.

A 72-year-old female patient with mixed rheumatic mitral valve disease and persistent atrial fibrillation underwent mitral valve replacement and suffered from a combined thrombosis of the bioprosthetic valve and the left atrium as soon as 2 days post operation. The patient immediately underwent repeated valve replacement and left atrial thrombectomy. Yet, four days later the patient died due to the recurrent prosthetic valve and left atrial thrombosis which both resulted in an extremely low cardiac output. In this patient's case, the thrombosis was notable for the resistance to anticoagulant therapy as well as for aggressive neutrophil infiltration and release of neutrophil extracellular traps (NETs) within the clot, as demonstrated by immunostaining. The reasons behind these phenomena remained unclear, as no signs of sepsis or contamination of the BHV were documented, although the patient was diagnosed with inherited thrombophilia that could impede the fibrinolysis. The described case highlights the hazard of immunothrombosis upon valve replacement and elucidates its mechanisms in this surgical setting.

Biomaterials Based on Carbon Nanotube Nanocomposites of Poly(styrene-b-isobutylene-b-styrene): The Effect of Nanotube Content on the Mechanical Properties, Biocompatibility and Hemocompatibility.

Nanocomposites based on poly(styrene-block-isobutylene-block-styrene) (SIBS) and single-walled carbon nanotubes (CNTs) were prepared and characterized in terms of tensile strength as well as bio- and hemocompatibility. It was shown that modification of CNTs using dodecylamine (DDA), featured by a long non-polar alkane chain, provided much better dispersion of nanotubes in SIBS as compared to unmodified CNTs. As a result of such modification, the tensile strength of the nanocomposite based on SIBS with low molecular weight (Mn = 40,000 g mol-1) containing 4% of functionalized CNTs was increased up to 5.51 ± 0.50 MPa in comparison with composites with unmodified CNTs (3.81 ± 0.11 MPa). However, the addition of CNTs had no significant effect on SIBS with high molecular weight (Mn~70,000 g mol-1) with ultimate tensile stress of pure polymer of 11.62 MPa and 14.45 MPa in case of its modification with 1 wt% of CNT-DDA. Enhanced biocompatibility of nanocomposites as compared to neat SIBS has been demonstrated in experiment with EA.hy 926 cells. However, the platelet aggregation observed at high CNT concentrations can cause thrombosis. Therefore, SIBS with higher molecular weight (Mn~70,000 g mol-1) reinforced by 1-2 wt% of CNTs is the most promising material for the development of cardiovascular implants such as heart valve prostheses.

Aortography Keypoint Tracking for Transcatheter Aortic Valve Implantation Based on Multi-Task Learning.

Currently, transcatheter aortic valve implantation (TAVI) represents the most efficient treatment option for patients with aortic stenosis, yet its clinical outcomes largely depend on the accuracy of valve positioning that is frequently complicated when routine imaging modalities are applied. Therefore, existing limitations of perioperative imaging underscore the need for the development of novel visual assistance systems enabling accurate procedures. In this paper, we propose an original multi-task learning-based algorithm for tracking the location of anatomical landmarks and labeling critical keypoints on both aortic valve and delivery system during TAVI. In order to optimize the speed and precision of labeling, we designed nine neural networks and then tested them to predict 11 keypoints of interest. These models were based on a variety of neural network architectures, namely MobileNet V2, ResNet V2, Inception V3, Inception ResNet V2 and EfficientNet B5. During training and validation, ResNet V2 and MobileNet V2 architectures showed the best prediction accuracy/time ratio, predicting keypoint labels and coordinates with 97/96% accuracy and 4.7/5.6% mean absolute error, respectively. Our study provides evidence that neural networks with these architectures are capable to perform real-time predictions of aortic valve and delivery system location, thereby contributing to the proper valve positioning during TAVI.

Real-time coronary artery stenosis detection based on modern neural networks.

Invasive coronary angiography remains the gold standard for diagnosing coronary artery disease, which may be complicated by both, patient-specific anatomy and image quality. Deep learning techniques aimed at detecting coronary artery stenoses may facilitate the diagnosis. However, previous studies have failed to achieve superior accuracy and performance for real-time labeling. Our study is aimed at confirming the feasibility of real-time coronary artery stenosis detection using deep learning methods. To reach this goal we trained and tested eight promising detectors based on different neural network architectures (MobileNet, ResNet-50, ResNet-101, Inception ResNet, NASNet) using clinical angiography data of 100 patients. Three neural networks have demonstrated superior results. The network based on Faster-RCNN Inception ResNet V2 is the most accurate and it achieved the mean Average Precision of 0.95, F1-score 0.96 and the slowest prediction rate of 3 fps on the validation subset. The relatively lightweight SSD MobileNet V2 network proved itself as the fastest one with a low mAP of 0.83, F1-score of 0.80 and a mean prediction rate of 38 fps. The model based on RFCN ResNet-101 V2 has demonstrated an optimal accuracy-to-speed ratio. Its mAP makes up 0.94, F1-score 0.96 while the prediction speed is 10 fps. The resultant performance-accuracy balance of the modern neural networks has confirmed the feasibility of real-time coronary artery stenosis detection supporting the decision-making process of the Heart Team interpreting coronary angiography findings.

Biocompatible Nanocomposites Based on Poly(styrene-block-isobutylene-block-styrene) and Carbon Nanotubes for Biomedical Application.

In this study, we incorporated carbon nanotubes (CNTs) into poly(styrene-block-isobutylene-block-styrene) (SIBS) to investigate the physical characteristics of the resulting nanocomposite and its cytotoxicity to endothelial cells. CNTs were dispersed in chloroform using sonication following the addition of a SIBS solution at different ratios. The resultant nanocomposite films were analyzed by X-ray microtomography, optical and scanning electron microscopy; tensile strength was examined by uniaxial tension testing; hydrophobicity was evaluated using a sessile drop technique; for cytotoxicity analysis, human umbilical vein endothelial cells were cultured on SIBS-CNTs for 3 days. We observed an uneven distribution of CNTs in the polymer matrix with sporadic bundles of interwoven nanotubes. Increasing the CNT content from 0 wt% to 8 wt% led to an increase in the tensile strength of SIBS films from 4.69 to 16.48 MPa. The engineering normal strain significantly decreased in 1 wt% SIBS-CNT films in comparison with the unmodified samples, whereas a further increase in the CNT content did not significantly affect this parameter. The incorporation of CNT into the SIBS matrix resulted in increased hydrophilicity, whereas no cytotoxicity towards endothelial cells was noted. We suggest that SIBS-CNT may become a promising material for the manufacture of implantable devices, such as cardiovascular patches or cusps of the polymer heart valve.

A New Nanocomposite Copolymer Based On Functionalised Graphene Oxide for Development of Heart Valves.

Polymeric heart valves seem to be an attractive alternative to mechanical and biological prostheses as they are more durable, due to the superior properties of novel polymers, and have the biocompatibility and hemodynamics comparable to tissue substitutes. This study reports a comprehensive assessment of a nanocomposite based on the functionalised graphene oxide and poly(carbonate-urea)urethane with the trade name "Hastalex" in comparison with GORE-TEX, a commercial polymer routinely used for cardiovascular medical devices. Experimental data have proved that GORE-TEX has a 2.5-fold (longitudinal direction) and 3.5-fold (transverse direction) lower ultimate tensile strength in comparison with Hastalex (p < 0.05). The contact angles of Hastalex surfaces (85.2 ± 1.1°) significantly (p < 0.05) are lower than those of GORE-TEX (127.1 ± 6.8°). The highest number of viable cells Ea.hy 926 is on the Hastalex surface exceeding 7.5-fold when compared with the GORE-TEX surface (p < 0.001). The platelet deformation index for GORE-TEX is 2-fold higher than that of Hastalex polymer (p < 0.05). Calcium content is greater for GORE-TEX (8.4 mg/g) in comparison with Hastalex (0.55 mg/g). The results of this study have proven that Hastalex meets the main standards required for manufacturing artificial heart valves and has superior mechanical, hemocompatibility and calcific resistance properties in comparison with GORE-TEX.