A New Image Registration Algorithm Based on Evidential Reasoning

Combining Sparse and Dense Features to Improve Multi-Modal Registration for Brain DTI Images

A new solution to overcome the constraints of multimodality medical intra-subject image registration is proposed, using the mutual information (MI) of image histogram-oriented gradients as a new matching criterion. We present a rigid, multi-modal image registration algorithm based on linear transformation and oriented gradients for the alignment of T2-weighted (T2w) images (as a fixed reference) and diffusion tensor imaging (DTI) (b-values of 500 and 1250 s/mm2) as floating images of three patients to compensate for the motion during the acquisition process. Diffusion MRI is very sensitive to motion, especially when the intensity and duration of the gradient pulses (characterized by the b-value) increases. The proposed method relies on the whole brain surface and addresses the variability of anatomical features into an image stack. The sparse features refer to corners detected using the Harris corner detector operator, while dense features use all image pixels through the image histogram of oriented gradients (HOG) as a measure of the degree of statistical dependence between a pair of registered images. HOG as a dense feature is focused on the structure and extracts the oriented gradient image in the x and y directions. MI is used as an objective function for the optimization process. The entropy functions and joint entropy function are determined using the HOGs data. To determine the best image transformation, the fiducial registration error (FRE) measure is used. We compare the results against the MI-based intensities results computed using a statistical intensity relationship between corresponding pixels in source and target images. Our approach, which is devoted to the whole brain, shows improved registration accuracy, robustness, and computational cost compared with the registration algorithms, which use anatomical features or regions of interest areas with specific neuroanatomy. Despite the supplementary HOG computation task, the computation time is comparable for MI-based intensities and MI-based HOG methods.

Using Image Registration and Machine Learning to Develop a Workstation Tool for Rapid Analysis of Glomeruli in Medical Renal Biopsies

Background

Prescreening of biopsies has the potential to improve pathologists' workflow. Tools that identify features and display results in a visually thoughtful manner can enhance efficiency, accuracy, and reproducibility. Machine learning for detection of glomeruli ensures comprehensive assessment and registration of four different stains allows for simultaneous navigation and viewing.

Methods

Medical renal core biopsies (4 stains each) were digitized using a Leica SCN400 at ×40 and loaded into the Corista Quantum research platform. Glomeruli were manually annotated by pathologists. The tissue on the 4 stains was registered using a combination of keypoint- and intensity-based algorithms, and a 4-panel simultaneous viewing display was created. Using a training cohort, machine learning convolutional neural net (CNN) models were created to identify glomeruli in all stains, and merged into composite fields of views (FOVs). The sensitivity and specificity of glomerulus detection, and FOV area for each detection were calculated.

Results

Forty-one biopsies were used for training (28) and same-batch evaluation (6). Seven additional biopsies from a temporally different batch were also evaluated. A variant of AlexNet CNN, used for object recognition, showed the best result for the detection of glomeruli with same-batch and different-batch evaluation: Same-batch sensitivity 92%, "modified" specificity 89%, average FOV size represented 0.8% of the total slide area; different-batch sensitivity 90%, "modified" specificity 98% and average FOV size 1.6% of the total slide area.

Conclusions

Glomerulus detection in the best CNN model shows that machine learning algorithms may be accurate for this task. The added benefit of biopsy registration with simultaneous display and navigation allows reviewers to move from one machine-generated FOV to the next in all 4 stains. Together these features could increase both efficiency and accuracy in the review process.

Effect of Image Fusion on Vegetation Index Quality—A Comparative Study from Gaofen-1, Gaofen-2, Gaofen-4, Landsat-8 OLI and MODIS Imagery

In recent years, the use of image fusion method has received increasing attention in remote sensing, vegetation cover changes, vegetation indices (VIs) mapping, etc. For making high-resolution and good quality (with low-cost) VI mapping from a fused image, its quality and underlying factors need to be identified properly. For example, same-sensor image fusion generally has a higher spatial resolution ratio (SRR) (1:3 to 1:5) but multi-sensor fusion has a lower SRR (1:8 to 1:10). In addition to SRR, there might be other factors affecting the fused vegetation index (FVI) result which have not been investigated in detail before. In this research, we used a strategy on image fusion and quality assessment to find the effect of image fusion for VI quality using Gaofen-1 (GF1), Gaofen-2 (GF2), Gaofen-4 (GF4), Landsat-8 OLI, and MODIS imagery with their panchromatic (PAN) and multispectral (MS) bands in low SRR (1:6 to 1:15). For this research, we acquired a total of nine images (4 PAN+5 MS) on the same (almost) date (GF1, GF2, GF4 and MODIS images were acquired on 2017/07/13 and the Landsat-8 OLI image was acquired on 2017/07/17). The results show that image fusion has the least impact on Green Normalized Vegetation Index (GNDVI) and Atmospherically Resistant Vegetation Index (ARVI) compared to other VIs. The quality of VI is mostly insensitive with image fusion except for the high-pass filter (HPF) algorithm. The subjective and objective quality evaluation shows that Gram-Schmidt (GS) fusion has the least impact on FVI quality, and with decreasing SRR, the FVI quality is decreasing at a slow rate. FVI quality varies with types image fusion algorithms and SRR along with spectral response function (SRF) and signal-to-noise ratio (SNR). However, the FVI quality seems good even for small SRR (1:6 to 1:15 or lower) as long as they have good SNR and minimum SRF effect. The findings of this study could be cost-effective and highly applicable for high-quality VI mapping even in small SRR (1:15 or even lower).

Non-Rigid Multi-Modal 3D Medical Image Registration Based on Foveated Modality Independent Neighborhood Descriptor

The non-rigid multi-modal three-dimensional (3D) medical image registration is highly challenging due to the difficulty in the construction of similarity measure and the solution of non-rigid transformation parameters. A novel structural representation based registration method is proposed to address these problems. Firstly, an improved modality independent neighborhood descriptor (MIND) that is based on the foveated nonlocal self-similarity is designed for the effective structural representations of 3D medical images to transform multi-modal image registration into mono-modal one. The sum of absolute differences between structural representations is computed as the similarity measure. Subsequently, the foveated MIND based spatial constraint is introduced into the Markov random field (MRF) optimization to reduce the number of transformation parameters and restrict the calculation of the energy function in the image region involving non-rigid deformation. Finally, the accurate and efficient 3D medical image registration is realized by minimizing the similarity measure based MRF energy function. Extensive experiments on 3D positron emission tomography (PET), computed tomography (CT), T1, T2, and (proton density) PD weighted magnetic resonance (MR) images with synthetic deformation demonstrate that the proposed method has higher computational efficiency and registration accuracy in terms of target registration error (TRE) than the registration methods that are based on the hybrid L-BFGS-B and cat swarm optimization (HLCSO), the sum of squared differences on entropy images, the MIND, and the self-similarity context (SSC) descriptor, except that it provides slightly bigger TRE than the HLCSO for CT-PET image registration. Experiments on real MR and ultrasound images with unknown deformation have also be done to demonstrate the practicality and superiority of the proposed method.

Eliminating the Effect of Image Border with Image Periodic Decomposition for Phase Correlation Based Remote Sensing Image Registration

In the remote sensing community, accurate image registration is the prerequisite of the subsequent application of remote sensing images. Phase correlation based image registration has drawn extensive attention due to its high accuracy and high efficiency. However, when the Discrete Fourier Transform (DFT) of an image is computed, the image is implicitly assumed to be periodic. In practical application, it is impossible to meet the periodic condition that opposite borders of an image are alike, and image always shows strong discontinuities across the frame border. The discontinuities cause a severe artifact in the Fourier Transform, namely the known cross structure composed of high energy coefficients along the axes. Here, this phenomenon was referred to as effect of image border. Even worse, the effect of image border corrupted its registration accuracy and success rate. Currently, the main solution is blurring out the border of the image by weighting window function on the reference and sensed image. However, the approach also inevitably filters out non-border information of an image. The existing understanding is that the design of window function should filter as little information as possible, which can improve the registration success rate and accuracy of methods based on phase correlation. In this paper, another approach of eliminating the effect of image border is proposed, namely decomposing the image into two images: one being the periodic image and the other the smooth image. Replacing the original image by the periodic one does not suffer from the effect on the image border when applying Fourier Transform. The smooth image is analogous to an error image, which has little information except at the border. Extensive experiments were carried out and showed that the novel algorithm of eliminating the image border can improve the success rate and accuracy of phase correlation based image registration in some certain cases. Additionally, we obtained a new understanding of the role of window function in eliminating the effect of image border, which is helpful for researchers to select the optimal method of eliminating the effect of image border to improve the registration success rate and accuracy.