Faculty Opinions recommendation of Spatial normalization of brain images with focal lesions using cost function masking.

Abstract With the ever-increasing need for concealing messages within cover media like image, video, and audio, numerous attempts have been developed for steganography. Most of the steganographic techniques perform their embedding operation on the cover image without selecting a better location. The right selection of location for embedding the information can lead to high imperceptibility and robustness. Accordingly, in this paper, we develop a new cost function for estimating the cost of every pixel to identify the good location to embed the message data. The proposed cost estimation procedure utilizes multiple parameters like wavelet coefficient, edge transformation, and pixel intensity. The proposed cost matrix is then utilized to embed the message data into the cover media using an embedding integer. The proposed steganographic technique is experimented with two magnetic resonance brain images, and the results are analyzed with the peak-to-peak signal-to-noise ratio (PSNR) and mean square error. The robustness analysis ensured that the proposed steganographic technique outperforms the existing methods by reaching the maximum PSNR of 72.74 dB.

Download Full-text

Automatic Landmark-guided Bijective Brain Image Registration by Composing Region-based Locally Diffeomorphic Warpings

10.1101/2020.04.16.045609 ◽

2020 ◽

Author(s):

Hengda He ◽

Qolamreza R. Razlighi

Keyword(s):

Human Brain ◽

Cognitive Aging ◽

Brain Morphology ◽

Spatial Normalization ◽

Weighted Interpolation ◽

Brain Images ◽

Whole Brain ◽

Normalization Methods ◽

Aging Populations ◽

Inverse Distance Weighted Interpolation

AbstractAs the size of the neuroimaging cohorts being increased to address key questions in the field of cognitive neuroscience, cognitive aging, and neurodegenerative diseases, the accuracy of the spatial normalization as an essential pre-processing step becomes extremely important in the neuroimaging processing pipeline. Existing spatial normalization methods have poor accuracy particularly when dealing with the highly convoluted human cerebral cortex and when brain morphology is severely altered (e.g. clinical and aging populations). To address this shortcoming, we propose to implement and evaluate a novel landmark-guided region-based spatial normalization technique that takes advantage of the existing surface-based human brain parcellation to automatically identify and match regional landmarks. To simplify the non-linear whole brain registration, the identified landmarks of each region and their counterparts are registered independently with large diffeomorphic (topology preserving) deformation via geodesic shooting. The regional diffeomorphic warping fields were combined by an inverse distance weighted interpolation technique to have a smooth global warping field for the whole brain. To ensure that the final warping field is diffeomorphic, we used simultaneously forward and reverse maps with certain symmetric constraints to yield bijectivity. We have evaluated our proposed method using both simulated and real (structural and functional) human brain images. Our evaluation shows that our method can enhance structural correspondence up to around 86%, a 67% improvement compared to the existing state-of-the-art method. Such improvement also increases the sensitivity and specificity of the functional imaging studies by about 17%, reducing the required number of subjects and subsequent costs. We conclude that our proposed method can effectively substitute existing substandard spatial normalization methods to deal with the demand of large cohorts and the need for investigating clinical and aging populations.

Download Full-text