Gradient Of Mutual Information Research Articles

Region-of-interest intra-arc MV imaging to facilitate sub-mm positional accuracy with minimal imaging dose during treatment deliveries of small cranial lesions.

To automate the generation of region-of-interest (ROI) apertures for use with megavoltage imaging for online positional corrections during cranial stereotactic radiosurgery. Digitally reconstructed radiographs (DRRs) were created for a 3D-printed skull phantom at 5degree gantry angle increments for a three-arc beam arrangement. At each angle, 3000 random rectangular apertures were generated, and 100 shifts on a grid were applied to the anatomy within the frame. For all shifts, the mutual information (MI) between the shifted and unshifted DRR was calculated to derive an average MI gradient. The top 10% of apertures that minimized registration errors were overlaid and discretely thresholded to generate imaging plans. Imaging was acquired with the skull while implementing simulated patient motion on a linac. Control point-specific couch motions were derived to align the skull to its planned positioning. Apertures with a range of repositioning errors less than 0.1mm possessed a 42% larger average MI gradient when compared with apertures with a range greater than 1mm. Dose calculations with Monte Carlo exhibited an 84% reduction in the dose received by 50% of the skull with the 50% thresholded plan when compared to a constant 22×22cm2 imaging plan. For all different imaging plans (with and without motion), the calculated median 3D-errors with respect to the tracking of a metal-BB fiducial positioned at isocenter in the skull were sub-mm except for the 80% thresholded plan. Sub-mm positional errors are achievable with couch motions derived from control point-specific ROI imaging. Smaller apertures that conform to an anatomical ROI can be utilized to minimize the imaging dose incurred at the expense of larger errors.

Multimodal non-rigid image registration based on elastodynamics

In this paper, we present a new multimodal image registration technique established on elastodynamics notion. The main idea behind this concept is the progression of waves on an elastic body as soon as it is disturbed from its initial rest state. We propose to solve the multimodal registration problem by modeling the non-linear deformations as elastic waves and iteratively solving the elastodynamics wave equation to estimate the transformation. The inertial force in elastodynamics model is computed as the gradient of mutual information which considers the statistical relationship between the intensities of the images acquired using different imaging modalities. We tested our method on T1–T2 weighted MR brain image pairs and MR-CT brain image pairs. The proposed registration technique was compared against a variant of demons method proposed for multimodal images. The registration results were analyzed by examining the overlay images and by computing the normalized mutual information. The qualitative and quantitative analysis proved that our proposed method registers the images better than the compared method.

Evaluation of GMI and PMI diffeomorphic-based demons algorithms for aligning PET and CT Images.

Fusion of anatomic information in computed tomography (CT) and functional information in F18‐FDG positron emission tomography (PET) is crucial for accurate differentiation of tumor from benign masses, designing radiotherapy treatment plan and staging of cancer. Although current PET and CT images can be acquired from combined F18‐FDG PET/CT scanner, the two acquisitions are scanned separately and take a long time, which may induce potential positional errors in global and local caused by respiratory motion or organ peristalsis. So registration (alignment) of whole‐body PET and CT images is a prerequisite for their meaningful fusion. The purpose of this study was to assess the performance of two multimodal registration algorithms for aligning PET and CT images. The proposed gradient of mutual information (GMI)‐based demons algorithm, which incorporated the GMI between two images as an external force to facilitate the alignment, was compared with the point‐wise mutual information (PMI) diffeomorphic‐based demons algorithm whose external force was modified by replacing the image intensity difference in diffeomorphic demons algorithm with the PMI to make it appropriate for multimodal image registration. Eight patients with esophageal cancer(s) were enrolled in this IRB‐approved study. Whole‐body PET and CT images were acquired from a combined F18‐FDG PET/CT scanner for each patient. The modified Hausdorff distance (dMH) was used to evaluate the registration accuracy of the two algorithms. Of all patients, the mean values and standard deviations (SDs) of dMH were 6.65 (± 1.90) voxels and 6.01 (± 1.90) after the GMI‐based demons and the PMI diffeomorphic‐based demons registration algorithms respectively. Preliminary results on oncological patients showed that the respiratory motion and organ peristalsis in PET/CT esophageal images could not be neglected, although a combined F18‐FDG PET/CT scanner was used for image acquisition. The PMI diffeomorphic‐based demons algorithm was more accurate than the GMI‐based demons algorithm in registering PET/CT esophageal images.PACS numbers: 87.57.nj, 87.57. Q‐, 87.57.uk

Playing Fetch with Your Robot: The Ability of Robots to Locate and Interact with Objects

The task addressed in this article is the localization of an unknown number of targets using a mobile robot equipped with a visual sensor. The estimation of the number of targets and their locations is done using a recursive Bayesian filter over random finite sets (RFSs), and the position of the robot is assumed to be known. We present a computationally tractable control law whereby the robot follows the gradient of mutual information between target locations and detections. The method is verified through real-world experimental trials, reliably detecting multiple targets and ignoring clutter obstacles.

A Registration Method for Multimodal Medical Images Using Contours Mutual Information

In recent years, mutual information has developed as a popular image registration measure especially in multimodality image registration. For different modality medical images, the contour of tissues or organs is similarity. In this paper, an effective new registration method of the multimodal medical images based on the contour mutual information is proposed. Firstly, get the contour through variational level set method. Secondly, within the contour pixels are assigned the same grayscale value, obtain two contour images. Finally, two contour images using mutual information as similarity measure for image registration. The experiment results show that the registration algorithm proposed in this paper can do more effectively and more accurately work than normalized mutual information and gradient mutual information.

Registration of PET and CT images based on multiresolution gradient of mutual information demons algorithm for positioning esophageal cancer patients

Accurate registration of 18F−FDG PET (positron emission tomography) and CT (computed tomography) images has important clinical significance in radiation oncology. PET and CT images are acquired from 18F−FDG PET/CT scanner, but the two acquisition processes are separate and take a long time. As a result, there are position errors in global and deformable errors in local caused by respiratory movement or organ peristalsis. The purpose of this work was to implement and validate a deformable CT to PET image registration method in esophageal cancer to eventually facilitate accurate positioning the tumor target on CT, and improve the accuracy of radiation therapy. Global registration was firstly utilized to preprocess position errors between PET and CT images, achieving the purpose of aligning these two images on the whole. Demons algorithm, based on optical flow field, has the features of fast process speed and high accuracy, and the gradient of mutual information‐based demons (GMI demons) algorithm adds an additional external force based on the gradient of mutual information (GMI) between two images, which is suitable for multimodality images registration. In this paper, GMI demons algorithm was used to achieve local deformable registration of PET and CT images, which can effectively reduce errors between internal organs. In addition, to speed up the registration process, maintain its robustness, and avoid the local extremum, multiresolution image pyramid structure was used before deformable registration. By quantitatively and qualitatively analyzing cases with esophageal cancer, the registration scheme proposed in this paper can improve registration accuracy and speed, which is helpful for precisely positioning tumor target and developing the radiation treatment planning in clinical radiation therapy application.PACS numbers: 87.57.nj, 87.57.Q‐, 87.57.uk

Postearthquake Building Damage Assessment Using Multi-Mutual Information From Pre-Event Optical Image and Postevent SAR Image

An approach of multi-mutual information (M-MI) is presented for change detection and evaluation of building damages after an earthquake. Fusion of very high resolution pre-event optical and postevent synthetic aperture radar (SAR) images becomes feasible for timely evaluation of earthquake losses. Based on the geometric parameters extracted from an optical pre-event image, SAR images of rectangular building objects, i.e., nondamaged or damaged, are first numerically simulated by our mapping and projection approach and are then, using M-MI, applied to similarity analysis with the real postevent SAR image. Three models of building damages, i.e., collapsed, subsided, and deformed, are proposed for classifying mutual information (MI). The M-MI, including normalized MI (NMI), gradient MI (GMI), and regional MI (RMI), are all applied and compared for MI change detection of building damages. Based on the maximum, mean value, and height deviation of NMI, GMI, and RMI, the building damages can be detected and evaluated. As an example, the Ikonos pre-event and GeoEye postevent optical images and COSMO-SkyMed and Radarsat-2 postevent SAR images during the 2010 Haiti earthquake are applied in this M-MI experiment.

Registration of 3D CT Data to 2D Endoscopic Image using a Gradient Mutual Information based Viewpoint Matching for Image-Guided Medialization Laryngoplasty

Registration of 3D CT Data to 2D Endoscopic Image using a Gradient Mutual Information based Viewpoint Matching for Image-Guided Medialization Laryngoplasty

Adaptive filtering under maximum mutual information criterion

The maximum mutual information (MaxMI) criterion is used as the adaptation cost for the adaptive filtering. This criterion is robust to measure distortions, and has strong connection with traditional mean-square error (MSE) criterion. Under Gaussian assumption, the closed-form solution of the finite impulse response (FIR) filter is obtained. Further, based on the kernel density estimation, the stochastic mutual information gradient (SMIG) algorithm is derived. Simulation results emphasize the robustness of this new algorithm.

Gradient of mutual information in linear vector Gaussian channels

This paper considers a general linear vector Gaussian channel with arbitrary signaling and pursues two closely related goals: i) closed-form expressions for the gradient of the mutual information with respect to arbitrary parameters of the system, and ii) fundamental connections between information theory and estimation theory. Generalizing the fundamental relationship recently unveiled by Guo, Shamai, and Verdu/spl acute/, we show that the gradient of the mutual information with respect to the channel matrix is equal to the product of the channel matrix and the error covariance matrix of the best estimate of the input given the output. Gradients and derivatives with respect to other parameters are then found via the differentiation chain rule.