Abstract. Computational methods are presented that can automatically detect the length and width of meibomian glands imaged by infrared meibography without requiring any input from the user. The images are then automatically classified. The length of the glands are detected by first normalizing the pixel intensity, extracting stationary points, and then applying morphological operations. Gland widths are detected using scale invariant feature transform and analyzed using Shannon entropy. Features based on the gland lengths and widths are then used to train a linear classifier to accurately differentiate between healthy (specificity 96.1%) and unhealthy (sensitivity 97.9%) meibography images. The user-free computational method is fast, does not suffer from inter-observer variability, and can be useful in clinical studies where large number of images needs to be analyzed efficiently.

We present a comparative ab initio study of surface diffusion of Li and Na on planar and curved graphene. The barrier for diffusion is~0.1 eV lower for Na than for Li, and is changed significantly by curvature. The maximum change is similar for Li for Na, of the order of ±0.1 eV on the convex and concave sides. The difference in barrier for metal atoms adsorbed on the concave and convex sides can reach 0.2 eV. This modulation of the diffusion barrier by curvature is therefore expected to affect significantly the rate capability of graphene-based anodes.

No abstract

The density of states of continuous models is known to span many orders of magnitudes at different energies due to the small volume of phase space near the ground state. Consequently, the traditional Wang-Landau sampling which uses the same trial move for all energies faces difficulties sampling the low entropic states. We developed an adaptive variant of the Wang-Landau algorithm that very effectively samples the density of states of continuous models across the entire energy range. By extending the acceptance ratio method of Bouzida, Kumar, and Swendsen such that the step size of the trial move and acceptance rate are adapted in an energy-dependent fashion, the random walker efficiently adapts its sampling according to the local phase space structure.The Wang-Landau modification factor is also made energy-dependent in accordance with the step size, enhancing the accumulation of the density of states. Numerical simulations show that our proposed method performs much better than the traditional Wang-Landau sampling.

We study the mechanism behind dynamical trappings experienced during Wang-Landau sampling of continuous systems reported by several authors. Trapping is caused by the random walker coming close to a local energy extremum, although the mechanism is different from that of critical slowing down encountered in conventional molecular dynamics or Monte Carlo simulations. When trapped, the random walker misses entire or even several stages of Wang-Landau modification factor reduction, leading to inadequate sampling of configuration space and a rough density-of-states even though the modification factor has been reduced to very small values. Trapping is dependent on specific systems, the choice of energy bins, and Monte Carlo step size, making it highly unpredictable. A general, simple, and effective solution is proposed where the configurations of multiple parallel Wang-Landau trajectories are inter-swapped to prevent trapping. We also explain why swapping frees the random walker from such traps. The efficacy of the proposed algorithm is demonstrated.