The localization of neural sources is a crucial issue in medical, scientific and technical applications. However, the dipole sources in the brain may vary in nature and change constantly with time, adding to the difficulty of source localization. In this paper, the dynamic neural sources in the brain are simulated by sequential Monte-Carlo (MC) method, based on electroencephalography (EEG) data. Firstly, the EEG data were considered as a state space model. Considering the nonlinearity of the EEG data, the sequential M-C method was introduced as a particle filter, and the Metropolis-Hastings (M-H) resampling was employed to alleviate the particle impoverishment of general particle filter. The accuracy of our method in source localization was verified through two experiments, using both synthetic and real data. The research results shed important new light on the research of brain neurology.

Tracking and detection of neural activity has numerous applications in the medical research field. By considering neural sources, it can be monitored by electroencephalography (EEG). In this paper, we focus primarily on developing advanced signal processing methods for locating neural sources. Due to its high performance in state estimation and tracking, particle filter was used to locate neural sources. However, particle degeneracy limits the performance of particle filters in the most utmost situations. A few resampling methods were subsequently proposed to ease this issue. These resampling methods, however, take on heavy computational costs. In this article, we aim to investigate the Partial Stratified Resampling algorithm which is time-efficient that can be used to locate neural sources and compare them to conventional resampling algorithms. This work is aimed at reflecting on the capabilities of various resampling algorithms and estimating the performance of locating neural sources. Simulated data and real EEG data are used to conduct evaluation and comparison experiments.

To recover the neural activity from Magnetoencephalography (MEG) and Electroencephalography (EEG) measurements, we need to solve the inverse problem by utilizing the relation between dipole sources and the data generated by dipolar sources. In this study, we propose a new approach based on the implementation of a particle filter (PF) that uses minimum sampling variance resampling methodology to track the neural dipole sources of cerebral activity. We use this approach for the EEG data and demonstrate that it can naturally estimate the sources more precisely than the traditional systematic resampling scheme in PFs.

Image segmentation in brain magnetic resonance imaging (MRI) largely relates to dividing brain tissue into components like white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF). Using the segmentation outputs, medical images can be 3D reconstructed and visualized efficiently. It is common for MRI pictures to have issues such as partial volume effects, asymmetrical grayscale, and noise. As a result, high accuracy in brain MRI picture segmentation is challenging to achieve in practical applications. In this paper, we developed an effective algorithm for brain MRI image segmentation utilizing a combination of statistical and partial differential equation-based approaches, based on a neuro-mechanical model. The findings of this work demonstrate that by combining various segmentation approaches, it is possible to quickly segment brain MRI data at a degree of precision necessary for different applications. Here, we show that when we use nonlinear filtering, [Formula: see text]-means clustering, and active contour modeling, we can get very good results when we segment brain MRI images. It is clear that the proposed approach has higher segmentation performance and can properly separate brain tissue from a large number of MRI images.

No abstract