Application of Opening and Closing Morphology in Deep Learning-Based Brain Image Registration

In order to improve the registration accuracy of brain magnetic resonance images(MRI),some deep learning registration methods use segmentation images for training model.How-ever,the segmentation values are constant for each label,which leads to the gradient variation con-centrating on the boundary.Thus,the dense deformation field(DDF)is gathered on the boundary and there even appears folding phenomenon.In order to fully leverage the label information,the morphological opening and closing information maps are introduced to enlarge the non-zero gradi-ent regions and improve the accuracy of DDF estimation.The opening information maps supervise the registration model to focus on smaller,narrow brain regions.The closing information maps supervise the registration model to pay more attention to the complex boundary region.Then,opening and closing morphology networks(OC_Net)are designed to automatically generate open-ing and closing information maps to realize the end-to-end training process.Finally,a new registra-tion architecture,VM_(seg+oc),is proposed by combining OC_Net and VoxelMorph.Experimental results show that the registration accuracy of VM_(seg+oc) is significantly improved on LPBA40 and OASIS1 datasets.Especially,VM_(seg+oc) can well improve registration accuracy in smaller brain regions and narrow regions.

Calculating collision probability for long-term satellite encounters through the reachable domain method

ABSTRACT Satellite encounters during close operations,such as rendezvous,formation,and cluster flights,are typical long-term encounters.The collision probability in such an encounter is a primary safety concern.In this study,a parametric method is proposed to compute the long-term collision probability for close satellite operations with initial state uncertainty.Random relative state errors resulting from system uncertainty lead to possible deviated trajectories with respect to the nominal one.To describe such a random event meaningfully,each deviated trajectory sample should be mapped to a unique and time-independent element in a random variable(RV)space.In this study,the RV space was identified as the transformed state space at a fixed initial time.The physical dimensions of both satellites were characterized by a combined hard-body sphere.Transforming the combined hard-body sphere into the RV space yielded a derived ellipsoid,which evolved over time and swept out a derived collision volume.The derived collision volume was solved using the reachable domain method.Finally,the collision probability was computed by integrating a probability density function over the derived collision volume.The results of the proposed method were compared with those of a nonparametric computation-intensive Monte Carlo method.The relative difference between the two results was found to be<0.6%,verifying the accuracy of the proposed method.

Perturbation Analysis of Moore–Penrose Quasi-linear Projection Generalized Inverse of Closed Linear Operators in Banach Spaces

In this paper, we investigate the perturbation problem for the Moore-Penrose bounded quasi-linear projection generalized inverses of a closed linear operaters in Banach space. By the method of the perturbation analysis of bounded quasi-linear operators, we obtain an explicit perturbation theorem and error estimates for the Moore-Penrose bounded quasi-linear generalized inverse of closed linear operator under the T-bounded perturbation, which not only extend some known results on the perturbation of the oblique projection generalized inverse of closed linear operators, but also extend some known results on the perturbation of the Moore-Penrose metric generalized inverse of bounded linear operators in Banach spaces.