The value of CT-based radiomics in predicting the prognosis of acute pancreatitis.

Purpose: Early judgment of the progress of acute pancreatitis (AP) and timely intervention are crucial to the prognosis of patients. The purpose of this study was to investigate the application value of CT-based radiomics of pancreatic parenchyma in predicting the prognosis of early AP. Materials and methods: This retrospective study enrolled 137 patients diagnosed with AP (95 cases in the progressive group and 42 cases in the non-progressive group) who underwent CT scans. Patients were randomly divided into a training set (n = 95) and a validation set (n = 42) in a ratio of 7: 3. The region of interest (ROI) was outlined along the inner edge of the pancreatic parenchyma manually, and the Modified CT Severity Index (MCTSI) was assessed. After resampling and normalizing the CT image, a total of 2,264 radiomics features were extracted from the ROI. The radiomics features were downscaled and filtered using minimum redundancy maximum correlation (mRMR) and the least absolute shrinkage and selection operator algorithm (LASSO) regression, in turn, and the more optimal subset of radiomics features was selected. In addition, the radiomics score (rad-score) was calculated for each patient by the LASSO method. Clinical data were also analyzed to predict the prognosis of AP. Three prediction models, including clinical model, radiomics model, and combined clinical-radiomics model, are constructed. The effectiveness of each model was evaluated using receiver operating characteristic (ROC) curve analysis. The DeLong test was employed to compare the differences between the ROC curves. The decision curve analysis (DCA) is used to assess the net benefit of the model. Results: The mRMR algorithm and LASSO regression were used to select 13 radiomics features with high values. The rad-score of each texture feature was calculated to fuse MCTSI to establish the radiomics model, and both the clinical model and clinical-radiomics model were established. The clinical-radiomics model showed the best performance, the AUC and 95% confidence interval, accuracy, sensitivity, and specificity of the clinical-radiomics model in the training set were 0.984 (0.964-1.000), 0.947, 0.955, and 0.931, respectively. In the validation set, they were 0.942 (0.870-1.000), 0.929, 0.966, and 0.846, respectively. The Delong test showed that the predictive efficacy of the clinical-radiomics model was higher than that of the clinical model (Z = 2.767, p = 0.005) and the radiomics model (Z = 2.033, p = 0.042) in the validation set. Decision curve analysis demonstrated higher net clinical benefit for the clinical-radiomics model. Conclusion: The pancreatic parenchymal CT clinical-radiomics model has high diagnostic efficacy in predicting the progression of early AP patients, which is significantly better than the clinical or radiomics model. The combined model can help identify and determine the progression trend of patients with AP and improve the prognosis and survival of patients as early as possible.

Analytical solution of a linear nonlocal Poisson-Boltzmann equation with multiple charges in a spherical solute region surrounded by a water spherical shell.

In this paper, an analytical solution of a linear nonlocal Poisson-Boltzmann equation (NPBE) test model with multiple charges in a spherical solute region surrounded by a water spherical shell is derived as a single series of Legendre polynomials and modified spherical Bessel functions. The classic Kirkwood ball model is then shown to be a special case of the NPBE test model so that its analytical solution is regained from a double series of associated Legendre polynomials (derived by Kirkwood in 1934) to a new single series of Legendre polynomials, sharply reducing its computational cost. As an application of these series solutions, a comparison study is done to demonstrate the differences between the Kirkwood and NPBE test models.

An improved Poisson-Nernst-Planck ion channel model and numerical studies on effects of boundary conditions, membrane charges, and bulk concentrations.

In this paper, an improved Poisson-Nernst-Planck ion channel (PNPic) model is presented, along with its effective finite element solver and software package for an ion channel protein in a solution of multiple ionic species. Numerical studies are then done on the effects of boundary value conditions, membrane charges, and bulk concentrations on electrostatics and ionic concentrations for an ion channel protein, a gramicidin A (gA), and five different ionic solvents with up to four species. Numerical results indicate that the cation selectivity property of gA occurs within a central portion of ion channel pore, insensitively to any change of boundary value condition, membrane charge, or bulk concentration. Moreover, a numerical scheme for computing the electric currents induced by ion transports across membrane via an ion channel pore is presented and implemented as a part of the PNPic finite element package. It is then applied to the calculation of current-voltage curves, well validating the PNPic model and finite element package by electric current experimental data.

A Rapid, Accurate and Machine-Agnostic Segmentation and Quantification Method for CT-Based COVID-19 Diagnosis.

COVID-19 has caused a global pandemic and become the most urgent threat to the entire world. Tremendous efforts and resources have been invested in developing diagnosis, prognosis and treatment strategies to combat the disease. Although nucleic acid detection has been mainly used as the gold standard to confirm this RNA virus-based disease, it has been shown that such a strategy has a high false negative rate, especially for patients in the early stage, and thus CT imaging has been applied as a major diagnostic modality in confirming positive COVID-19. Despite the various, urgent advances in developing artificial intelligence (AI)-based computer-aided systems for CT-based COVID-19 diagnosis, most of the existing methods can only perform classification, whereas the state-of-the-art segmentation method requires a high level of human intervention. In this paper, we propose a fully-automatic, rapid, accurate, and machine-agnostic method that can segment and quantify the infection regions on CT scans from different sources. Our method is founded upon two innovations: 1) the first CT scan simulator for COVID-19, by fitting the dynamic change of real patients' data measured at different time points, which greatly alleviates the data scarcity issue; and 2) a novel deep learning algorithm to solve the large-scene-small-object problem, which decomposes the 3D segmentation problem into three 2D ones, and thus reduces the model complexity by an order of magnitude and, at the same time, significantly improves the segmentation accuracy. Comprehensive experimental results over multi-country, multi-hospital, and multi-machine datasets demonstrate the superior performance of our method over the existing ones and suggest its important application value in combating the disease.

A size-modified poisson-boltzmann ion channel model in a solvent of multiple ionic species: Application to voltage-dependent anion channel.

We present a new size-modified Poisson-Boltzmann ion channel (SMPBIC) model and use it to calculate the electrostatic potential, ionic concentrations, and electrostatic solvation free energy for a voltage-dependent anion channel (VDAC) on a biological membrane in a solution mixture of multiple ionic species. In particular, the new SMPBIC model adopts a membrane surface charge density and a natural Neumann boundary condition to reflect the charge effect of the membrane on the electrostatics of VDAC. To avoid the singularity difficulties caused by the atomic charges of VDAC, the new SMPBIC model is split into three submodels such that the solution of one of the submodels is obtained analytically and contains all the singularity points of the SMPBIC model. The other two submodels are then solved numerically much more efficiently than the original SMPBIC model. As an application of this SMPBIC submodel partitioning scheme, we derive a new formula for computing the electrostatic solvation free energy. Numerical results for a human VDAC isoform 1 (hVDAC1) in three different salt solutions, each with up to five different ionic species, confirm the significant effects of membrane surface charges on both the electrostatics and ionic concentrations. The results also show that the new SMPBIC model can describe well the anion selectivity property of hVDAC1, and that the new electrostatic solvation free energy formula can significantly improve the accuracy of the currently used formula. © 2019 Wiley Periodicals, Inc.

Integrated Computational Model of Lung Tissue Bioenergetics.

Altered lung tissue bioenergetics plays a key role in the pathogenesis of lung diseases. A wealth of information exists regarding the bioenergetic processes in mitochondria isolated from rat lungs, cultured pulmonary endothelial cells, and intact rat lungs under physiological and pathophysiological conditions. However, the interdependence of those processes makes it difficult to quantify the impact of a change in a single or multiple process(es) on overall lung tissue bioenergetics. Integrated computational modeling provides a mechanistic and quantitative framework for the bioenergetic data at different levels of biological organization. The objective of this study was to develop and validate an integrated computational model of lung bioenergetics using existing experimental data from isolated perfused rat lungs. The model expands our recently developed computational model of the bioenergetics of mitochondria isolated from rat lungs by accounting for glucose uptake and phosphorylation, glycolysis, and the pentose phosphate pathway. For the mitochondrial region of the model, values of kinetic parameters were fixed at those estimated in our recent model of the bioenergetics of mitochondria isolated from rat lungs. For the cytosolic region of the model, intrinsic parameters such as apparent Michaelis constants were determined based on previously published enzyme kinetics data, whereas extrinsic parameters such as maximal reaction and transport velocities were estimated by fitting the model solution to published data from isolated rat lungs. The model was then validated by assessing its ability to predict existing experimental data not used for parameter estimation, including relationships between lung nucleotides content, lung lactate production rate, and lung energy charge under different experimental conditions. In addition, the model was used to gain novel insights on how lung tissue glycolytic rate is regulated by exogenous substrates such as glucose and lactate, and assess differences in the bioenergetics of mitochondria isolated from lung tissue and those of mitochondria in intact lungs. To the best of our knowledge, this is the first model of lung tissue bioenergetics. The model provides a mechanistic and quantitative framework for integrating available lung tissue bioenergetics data, and for testing novel hypotheses regarding the role of different cytosolic and mitochondrial processes in lung tissue bioenergetics.

An accelerated nonlocal Poisson-Boltzmann equation solver for electrostatics of biomolecule.

The nonlocal modified Poisson-Boltzmann equation (NMPBE) is one important variant of a commonly used dielectric continuum model, the Poisson-Boltzmann equation (PBE), for computing electrostatics of biomolecules. In this paper, an accelerated NMPBE solver is constructed by finite element and finite difference hybrid techniques. It is then programmed as a software package for computing electrostatic solvation and binding free energies for a protein in a symmetric 1:1 ionic solvent. Numerical results validate the new solver and its numerical stability. They also demonstrate that the new solver has much better performance than the corresponding finite element solver in terms of computer CPU time. Furthermore, they show that the binding free energies produced by NMPBE can match chemical experiment data better than those by PBE.

Nonlocal Poisson-Fermi double-layer models: Effects of nonuniform ion sizes on double-layer structure.

This paper reports a nonuniform ionic size nonlocal Poisson-Fermi double-layer model (nuNPF) and a uniform ionic size nonlocal Poisson-Fermi double-layer model (uNPF) for an electrolyte mixture of multiple ionic species, variable voltages on electrodes, and variable induced charges on boundary segments. The finite element solvers of nuNPF and uNPF are developed and applied to typical double-layer tests defined on a rectangular box, a hollow sphere, and a hollow rectangle with a charged post. Numerical results show that nuNPF can significantly improve the quality of the ionic concentrations and electric fields generated from uNPF, implying that the effect of nonuniform ion sizes is a key consideration in modeling the double-layer structure.

SMPBS: Web server for computing biomolecular electrostatics using finite element solvers of size modified Poisson-Boltzmann equation.

SMPBS (Size Modified Poisson-Boltzmann Solvers) is a web server for computing biomolecular electrostatics using finite element solvers of the size modified Poisson-Boltzmann equation (SMPBE). SMPBE not only reflects ionic size effects but also includes the classic Poisson-Boltzmann equation (PBE) as a special case. Thus, its web server is expected to have a broader range of applications than a PBE web server. SMPBS is designed with a dynamic, mobile-friendly user interface, and features easily accessible help text, asynchronous data submission, and an interactive, hardware-accelerated molecular visualization viewer based on the 3Dmol.js library. In particular, the viewer allows computed electrostatics to be directly mapped onto an irregular triangular mesh of a molecular surface. Due to this functionality and the fast SMPBE finite element solvers, the web server is very efficient in the calculation and visualization of electrostatics. In addition, SMPBE is reconstructed using a new objective electrostatic free energy, clearly showing that the electrostatics and ionic concentrations predicted by SMPBE are optimal in the sense of minimizing the objective electrostatic free energy. SMPBS is available at the URL: smpbs.math.uwm.edu © 2017 Wiley Periodicals, Inc.

Nonlocal Poisson-Fermi model for ionic solvent.

We propose a nonlocal Poisson-Fermi model for ionic solvent that includes ion size effects and polarization correlations among water molecules in the calculation of electrostatic potential. It includes the previous Poisson-Fermi models as special cases, and its solution is the convolution of a solution of the corresponding nonlocal Poisson dielectric model with a Yukawa-like kernel function. The Fermi distribution is shown to be a set of optimal ionic concentration functions in the sense of minimizing an electrostatic potential free energy. Numerical results are reported to show the difference between a Poisson-Fermi solution and a corresponding Poisson solution.

Analytical solutions of nonlocal Poisson dielectric models with multiple point charges inside a dielectric sphere.

The nonlocal dielectric approach has led to new models and solvers for predicting electrostatics of proteins (or other biomolecules), but how to validate and compare them remains a challenge. To promote such a study, in this paper, two typical nonlocal dielectric models are revisited. Their analytical solutions are then found in the expressions of simple series for a dielectric sphere containing any number of point charges. As a special case, the analytical solution of the corresponding Poisson dielectric model is also derived in simple series, which significantly improves the well known Kirkwood's double series expansion. Furthermore, a convolution of one nonlocal dielectric solution with a commonly used nonlocal kernel function is obtained, along with the reaction parts of these local and nonlocal solutions. To turn these new series solutions into a valuable research tool, they are programed as a free fortran software package, which can input point charge data directly from a protein data bank file. Consequently, different validation tests can be quickly done on different proteins. Finally, a test example for a protein with 488 atomic charges is reported to demonstrate the differences between the local and nonlocal models as well as the importance of using the reaction parts to develop local and nonlocal dielectric solvers.

Hematoma volume affects the accuracy of computed tomographic angiography 'spot sign' in predicting hematoma expansion after acute intracerebral hemorrhage.

BACKGROUND: We hypothesized that the presence of tiny, enhancing foci ('spot sign') within acute hematomas is associated with hematoma expansion. METHODS: We retrospectively analyzed the effect of hematoma volume on accuracy of computed tomographic angiography (CTA) in predicting hematoma expansion in 312 patients with acute intracerebral hemorrhage (ICH). The patients were divided into 2 groups according to their initial hematoma volume (<30 vs. ≥30 ml). CTA was performed at admission and 24 h after initial presentation. RESULTS: The <30-ml group consisted of 203 patients of whom 42 had hematoma expansion (20.9%). The ≥30-ml group consisted of 109 patients of whom 34 had hematoma expansion (31.19%). In the <30-ml group, the sensitivity and specificity of CTA in predicting hematoma expansion were 71.4 and 93.8%, respectively. In the ≥30-ml group, the sensitivity and specificity of CTA were 85.7 and 91.9%, respectively. For all 312 patients, the area under the curve was 0.86 (p < 0.001, 95% CI 0.80-0.92); the sensitivity and specificity of CTA were 77.9 and 93.2%, respectively. CONCLUSIONS: CTA can reliably predict hematoma expansion in clinical practice, especially for hematomas >30 ml.

A molecular dynamics model of the Bt toxin Cyt1A and its validation by resonance energy transfer.

Cyt1A is a cytolytic toxin from Bacillus thuringiensis var. israelensis. A computer model of the toxin in solution was generated and validated by resonance energy transfer (RET). The average distance between the two tryptophans (residues 158 and 161) and the fluorescently labeled cysteine 190 was 2.16 nm, which closely matched the distance predicted in computer simulations, 2.2 nm. The simulation results were able to explain two previous experimental observations: (i) amino-acid sequences of all Cyt toxins contain four blocks of highly conserved residues; and (ii) several single-point mutations drastically abrogated Cyt1A's toxicity. Selective randomization of atomic coordinates in the computer model revealed that the conserved blocks are important for proper folding and stability of the toxin molecule. Replacing lysine 225 with alanine, a mutation that renders the toxin inactive, was shown to result in breaking the hydrogen bonds between K225 and V126, L123, and Y189. Calculated Helmholtz free energy difference of the inactive mutation K225A was higher by 12 kcal/mol and 5 kcal/mol than the values for the benign mutations K118A and K198A, respectively, which indicates that the K225A mutant is significantly destabilized. The normal-mode and principal-component analyses revealed that in the wild-type Cyt1A the region around the residue K225 is quite stationary, due to the hydrogen-bond network around K225. In contrast, pronounced twisting and stretching were observed in the mutant K225A, and the region around the residue K225 becomes unstable. Our results indicate that conformational differences in this mutant spread far away from the site of the mutation, suggesting that the mutant is inactivated due to an overall change in conformation and diminished stability rather than due to a localized alteration of a "binding" or "active" site.

An Improved Algorithm and Its Parallel Implementation for Solving a General Blood-Tissue Transport and Metabolism Model.

Fast algorithms for simulating mathematical models of coupled blood-tissue transport and metabolism are critical for the analysis of data on transport and reaction in tissues. Here, by combining the method of characteristics with the standard grid discretization technique, a novel algorithm is introduced for solving a general blood-tissue transport and metabolism model governed by a large system of one-dimensional semilinear first order partial differential equations. The key part of the algorithm is to approximate the model as a group of independent ordinary differential equation (ODE) systems such that each ODE system has the same size as the model and can be integrated independently. Thus the method can be easily implemented in parallel on a large scale multiprocessor computer. The accuracy of the algorithm is demonstrated for solving a simple blood-tissue exchange model introduced by Sangren and Sheppard (Bull. Math. Biophys. 15:387-394, 1953), which has an analytical solution. Numerical experiments made on a distributed-memory parallel computer (an HP Linux cluster) and a shared-memory parallel computer (a SGI Origin 2000) demonstrate the parallel efficiency of the algorithm.

An incomplete Hessian Newton minimization method and its application in a chemical database problem.

To efficiently solve a large scale unconstrained minimization problem with a dense Hessian matrix, this paper proposes to use an incomplete Hessian matrix to define a new modified Newton method, called the incomplete Hessian Newton method (IHN). A theoretical analysis shows that IHN is convergent globally, and has a linear rate of convergence with a properly selected symmetric, positive definite incomplete Hessian matrix. It also shows that the Wolfe conditions hold in IHN with a line search step length of one. As an important application, an effective IHN and a modified IHN, called the truncated-IHN method (T-IHN), are constructed for solving a large scale chemical database optimal projection mapping problem. T-IHN is shown to work well even with indefinite incomplete Hessian matrices. Numerical results confirm the theoretical results of IHN, and demonstrate the promising potential of T-IHN as an efficient minimization algorithm.

[Value of multi-slice computed tomography in diagnosis of coronary plaque characterization].

OBJECTIVE: To investigate the value of multi-slice computed tomography in diagnosis of different types of coronary atherosclerotic plaques. METHODS: Twenty-eight patients undergoing CT angiography (CTA) with normal coronary arteries were randomly selected to measure the CT values of different sections of the 4 main branches of coronary artery. Twenty-five specimens of human heart from the bodies of the patients who died of non-cardiogenic diseases were scanned by 16-slice CT scanner and 64-slice CT scanner: a mixture of CT contrast media and normal saline was injected into the coronary arteries to achieve in-vivo-like contrast enhancement within the coronary artery lumen to detect atherosclerotic plaques. The CT values of plaques were measured in several regions of interest (ROI) selected in each plaque. The CT images thus obtained were evaluated by 2 experienced radiologists. There are nine specimens with coronary atherosclerotic plaques among them. Then the atherosclerotic lesions in the coronary were made into tissue specimens to undergo pathological examination. RESULTS: 7560 CT values were obtained from the 28 patients. Thirty-eight atherosclerotic plaques were found by CAT in 9 heart specimens and confirmed by pathology. When the CT value of coronary lumen was 370 HU, the predominant lipid-rich plaque showed a mean CT value of 53 +/- 12 HU; the fibrous-rich plaque showed a mean CT value of 106 +/- 17 HU; and the calcified plaque showed a mean CT value of 429 +/- 94 HU measured by 16-slice CT; and the predominant lipid-rich plaque showed a mean CT value of 51 +/- 13 HU; the fibrous-rich plaque showed a mean CT value of 110 +/- 19 HU; and the calcified plaque showed a mean CT value of 435 +/- 87 HU measured by 64-slice CT. The CT value of the fibrous-rich plaque was significantly higher than that of the lipid-rich plaque (P = 0.008), and lower than that of the calcified plaque (P < 0.01). There was no significant difference between the results obtained by the two kinds of CT scanners. CONCLUSION: CTA can non-invasively assess the atherosclerotic plaques.

Molecular dynamics simulations of the cytolytic toxin Cyt1A in solution.

Cytolytic toxin Cyt1A from Bacillus thuringiensis var. israelensis is used as an environmentally friendly insecticide, but its mode of action has not been clearly established. One main obstacle seems to be the lack of the experimentally determined structure of the toxin. As a first step in computer simulations of Cyt1A, in this paper, a three-dimensional molecular structure of Cyt1A in solution was generated by homology modeling, potential energy minimization and molecular dynamics. Regions of the toxin molecule that manifest increased conformational flexibility--and thus are likely to participate in the initial membrane binding and conformational changes--were then identified. Finally, the simulated structure was used to study the effect of a single amino-acid mutation that is known to abrogate the toxicity of Cyt1A in vivo.