Artificial intelligence velocimetry reveals in vivo flow rates, pressure gradients, and shear stresses in murine perivascular flows.

Quantifying the flow of cerebrospinal fluid (CSF) is crucial for understanding brain waste clearance and nutrient delivery, as well as edema in pathological conditions such as stroke. However, existing in vivo techniques are limited to sparse velocity measurements in pial perivascular spaces (PVSs) or low-resolution measurements from brain-wide imaging. Additionally, volume flow rate, pressure, and shear stress variation in PVSs are essentially impossible to measure in vivo. Here, we show that artificial intelligence velocimetry (AIV) can integrate sparse velocity measurements with physics-informed neural networks to quantify CSF flow in PVSs. With AIV, we infer three-dimensional (3D), high-resolution velocity, pressure, and shear stress. Validation comes from training with 70% of PTV measurements and demonstrating close agreement with the remaining 30%. A sensitivity analysis on the AIV inputs shows that the uncertainty in AIV inferred quantities due to uncertainties in the PVS boundary locations inherent to in vivo imaging is less than 30%, and the uncertainty from the neural net initialization is less than 1%. In PVSs of N = 4 wild-type mice we find mean flow speed 16.33 ± 11.09 µm/s, volume flow rate 2.22 ± 1.983 × 103 µm3/s, axial pressure gradient ( - 2.75 ± 2.01)×10-4 Pa/µm (-2.07 ± 1.51 mmHg/m), and wall shear stress (3.00 ± 1.45)×10-3 Pa (all mean ± SE). Pressure gradients, flow rates, and resistances agree with prior predictions. AIV infers in vivo PVS flows in remarkable detail, which will improve fluid dynamic models and potentially clarify how CSF flow changes with aging, Alzheimer's disease, and small vessel disease.

A three-dimensional phase-field model for multiscale modeling of thrombus biomechanics in blood vessels.

Mechanical interactions between flowing and coagulated blood (thrombus) are crucial in dictating the deformation and remodeling of a thrombus after its formation in hemostasis. We propose a fully-Eulerian, three-dimensional, phase-field model of thrombus that is calibrated with existing in vitro experimental data. This phase-field model considers spatial variations in permeability and material properties within a single unified mathematical framework derived from an energy perspective, thereby allowing us to study effects of thrombus microstructure and properties on its deformation and possible release of emboli under different hemodynamic conditions. Moreover, we combine this proposed thrombus model with a particle-based model which simulates the initiation of the thrombus. The volume fraction of a thrombus obtained from the particle simulation is mapped to an input variable in the proposed phase-field thrombus model. The present work is thus the first computational study to integrate the initiation of a thrombus through platelet aggregation with its subsequent viscoelastic responses to various shear flows. This framework can be informed by clinical data and potentially be used to predict the risk of diverse thromboembolic events under physiological and pathological conditions.

Non-invasive Inference of Thrombus Material Properties with Physics-Informed Neural Networks.

We employ physics-informed neural networks (PINNs) to infer properties of biological materials using synthetic data. In particular, we successfully apply PINNs on inferring permeability and viscoelastic modulus from thrombus deformation data, which can be described by the fourth-order Cahn-Hilliard and Navier-Stokes Equations. In PINNs, the partial differential equations are encoded into a loss function, where partial derivatives can be obtained through automatic differentiation (AD). In addition to tackling the challenge of calculating the fourth-order derivative in the Cahn-Hilliard equation with AD, we introduce an auxiliary network along with the main neural network to approximate the second-derivative of the energy potential term. Our model can simultaneously predict unknown material parameters and velocity, pressure, and deformation gradient fields by merely training with partial information among all data, i.e., phase field and pressure measurements, while remaining highly flexible in sampling within the spatio-temporal domain for data acquisition. We validate our model by numerical solutions from the spectral/hp element method (SEM) and demonstrate its robustness by training it with noisy measurements. Our results show that PINNs can infer the material properties from noisy synthetic data, and thus they have great potential for inferring these properties from experimental multi-modality and multi-fidelity data.

Mechanisms underlying the efficacy and limitation of dopa and tetrahydrobiopterin therapies for the deficiency of GTP cyclohydrolase 1 revealed in a novel mouse model.

Dopa and tetrahydrobiopterin (BH4) supplementation are recommended therapies for the dopa-responsive dystonia caused by GTP cyclohydrolase 1 (GCH1, also known as GTPCH) deficits. However, the efficacy and mechanisms of these therapies have not been intensively studied yet. In this study, we tested the efficacy of dopa and BH4 therapies by using a novel GTPCH deficiency mouse model, Gch1KI/KI, which manifested infancy-onset motor deficits and growth retardation similar to the patients. First, dopa supplementation supported Gch1KI/KI mouse survival to adulthood, but residual motor deficits and dwarfism remained. Interestingly, RNAseq analysis indicated that while the genes participating in BH4 biosynthesis and regeneration were significantly increased in the liver, no significant changes were observed in the brain. Second, BH4 supplementation alone restored the growth of Gch1KI/KI pups only in early postnatal developmental stage. High doses of BH4 supplementation indeed restored the total brain BH4 levels, but brain dopamine deficiency remained. While total brain TH levels were relatively increased in the BH4 treated Gch1KI/KI mice, the TH in the striatum were still almost undetectable, suggesting differential BH4 requirements among brain regions. Last, the growth of Gch1KI/KI mice under combined therapy outperformed dopa or BH4 therapy alone. Notably, dopamine was abnormally high in more than half, but not all, of the treated Gch1KI/KI mice, suggesting the existence of variable synergetic effects of dopa and BH4 supplementation. Our results provide not only experimental evidence but also novel mechanistic insights into the efficacy and limitations of dopa and BH4 therapies for GTPCH deficiency.

Iron Removal from Metallurgical Grade Silicon Melts Using Synthetic Slags and Oxygen Injection.

Novel SiO2-CaO-CaF2-R2O-MgO based synthetic slags (R2O represents alkali metal oxides) with varied binary basicity values were used with oxygen injection to refine silicon melts and remove Fe from metallurgical-grade silicon. Silicon samples and slags at the silicon-slag interfaces were obtained during refinement. The compositions of the silicon samples were analyzed, and the quenched slag samples and mild cooling slags from the final crucible were inspected using scanning electron microscopy and energy dispersive X-ray spectroscopy. After 15 min of refinement, the Fe removal rate ranged from 52.3 to 60.1 wt%. During the refining process, the Fe-concentrated phase formed within the silicon droplets and was then transferred to the silicon-slag interfaces and wetted with slags. The Fe-concentrated phase at the silicon-slag interface can dissolve directly in the slags. It can also be transferred into the slag phase in the form of droplets, which can be affected by the binary basicity of the slags. Ti removal demonstrated a similar mechanism. Fe-bearing crystals were not detected in the quenched slag samples obtained during refinement, while complex Fe-bearing phases were detected in the final slag. This study demonstrates Fe removal from metallurgical-grade Si using slag refining methods and reveals the removal mechanism during the refinement.

Identifiability and predictability of integer- and fractional-order epidemiological models using physics-informed neural networks.

We analyze a plurality of epidemiological models through the lens of physics-informed neural networks (PINNs) that enable us to identify time-dependent parameters and data-driven fractional differential operators. In particular, we consider several variations of the classical susceptible-infectious-removed (SIR) model by introducing more compartments and fractional-order and time-delay models. We report the results for the spread of COVID-19 in New York City, Rhode Island and Michigan states and Italy, by simultaneously inferring the unknown parameters and the unobserved dynamics. For integer-order and time-delay models, we fit the available data by identifying time-dependent parameters, which are represented by neural networks. In contrast, for fractional differential models, we fit the data by determining different time-dependent derivative orders for each compartment, which we represent by neural networks. We investigate the structural and practical identifiability of these unknown functions for different datasets, and quantify the uncertainty associated with neural networks and with control measures in forecasting the pandemic.

Predictive modelling of thrombus formation in diabetic retinal microaneurysms.

Microaneurysms (MAs) are one of the earliest clinically visible signs of diabetic retinopathy (DR). Vision can be reduced at any stage of DR by MAs, which may enlarge, rupture and leak fluid into the neural retina. Recent advances in ophthalmic imaging techniques enable reconstruction of the geometries of MAs and quantification of the corresponding haemodynamic metrics, such as shear rate and wall shear stress, but there is lack of computational models that can predict thrombus formation in individual MAs. In this study, we couple a particle model to a continuum model to simulate the platelet aggregation in MAs with different shapes. Our simulation results show that under a physiologically relevant blood flow rate, thrombosis is more pronounced in saccular-shaped MAs than fusiform-shaped MAs, in agreement with recent clinical findings. Our model predictions of the size and shape of the thrombi in MAs are consistent with experimental observations, suggesting that our model is capable of predicting the formation of thrombus for newly detected MAs. This is the first quantitative study of thrombosis in MAs through simulating platelet aggregation, and our results suggest that computational models can be used to predict initiation and development of intraluminal thrombus in MAs as well as provide insights into their role in the pathophysiology of DR.

[Activity and transcriptional regulatory elements of the promoter in Arctic fox (Vulpes lagopus) ß-defensin103 gene].

The aim of this study was to screen the active regions and transcription factor binding sites in the promoter of the CBD103 gene related to Arctic fox coat color, and to provide a basis for revealing the molecular genetic mechanism of CBD103 gene regulating the coat color formation. The 5'-flanking region fragment 2 123 bp of Arctic fox CBD103 gene was cloned, and 4 truncated promoter reporter vectors of different lengths were constructed. The promoter activity was detected by the dual-luciferase reporter assay system. Point mutations were performed on the 3 predicted specificity protein 1 (Sp1) transcription factor binding sites in the highest promoter active region, and 3 mutant vectors were constructed. The activity was then detected by the dual-luciferase reporter assay system. The results showed that the region 1 656 (-1 604/+51) had the highest activity in the 4 truncated promoters of different lengths, and the promoter activity of the three mutant vectors constructed in this region were significantly lower than that of the wild type (fragment 1 656). The region of -1 604 /+51 was the core promoter region of CBD103 gene in Arctic fox and -1 552/-1 564, -1 439/-1 454 and -329/-339 regions were positive regulatory regions. This study successfully obtained the core promoter region and positive regulation regions of the Arctic fox CBD103 gene, which laid a foundation for further study on the molecular genetic mechanism of this gene regulating Arctic fox coat color.

Future of lignite resources: a life cycle analysis.

Lignite is a low-quality energy source which accounts for 13 % of China's coal reserves. It is imperative to improve the quality of lignite for large-scale utilization. To further explore and analyze the influence of various key processes on the environment and economic costs, a lignite drying and compression technology is evaluated using an integrated approach of life cycle assessment and life cycle costs. Results showed that lignite mining, direct air emissions, and electricity consumption have most significant impacts on the environment. An integrated evaluation of life cycle assessment and life cycle costs showed that the most significant contributor to the environmental impacts and economic costs was the lignite mining process. The impact of transportation and wastewater treatment process on the environment and economic costs was small enough to be ignored. Critical factors were identified for reducing the environmental and economic impacts of lignite drying and compression technology. These findings provide useful inputs for both industrial practice and policy making for exploitation, processing, and utilization of lignite resources.