# American Institute of Mathematical Sciences

• Previous Article
Existence and nonexistence of entire positive radial solutions for a class of Schrödinger elliptic systems involving a nonlinear operator
• DCDS-S Home
• This Issue
• Next Article
A direct method of moving planes for fully nonlinear nonlocal operators and applications

## Signed-distance function based non-rigid registration of image series with varying image intensity

 a. Department of Mathematics, FNSPE, Czech Technical University in Prague, Trojanova 13,120 00 Prague, Czech Republic b. Department of Radiology, Institute for clinical and experimental medicine, Vídeňská 1958/9, Praha 4,140 21, Czech Republic c. Inria, France d. LMS, Ecole Polytechnique, CNRS, Institut Polytechnique de Paris, France e. School of Biomedical Engineering & Imaging Sciences, St Thomas' Hospital, King's College London, UK

* Corresponding author: katerina.skardova@fjfi.cvut.cz

Received  January 2019 Revised  February 2020 Published  June 2020

In this paper we propose a method for locally adjusted optical flow-based registration of multimodal images, which uses the segmentation of object of interest and its representation by the signed-distance function (OF$^{dist}$ method). We deal with non-rigid registration of the image series acquired by the Modiffied Look-Locker Inversion Recovery (MOLLI) magnetic resonance imaging sequence, which is used for a pixel-wise estimation of $T_1$ relaxation time. The spatial registration of the images within the series is necessary to compensate the patient's imperfect breath-holding. The evolution of intensities and a large variation of image contrast within the MOLLI image series, together with the myocardium of left ventricle (the object of interest) typically not being the most distinct object in the scene, makes the registration challenging. The paper describes all components of the proposed OF$^{dist}$ method and their implementation. The method is then compared to the performance of a standard mutual information maximization-based registration method, applied either to the original image (MIM) or to the signed-distance function (MIM$^{dist}$). Several experiments with synthetic and real MOLLI images are carried out. On synthetic image with a single object, MIM performed the best, while OF$^{dist}$ and MIM$^{dist}$ provided better results on synthetic images with more than one object and on real images. When applied to signed-distance function of two objects of interest, MIM$^{dist}$ provided a larger registration error, but more homogeneously distributed, compared to OF$^{dist}$. For the real MOLLI image series with left ventricle pre-segmented using a level-set method, the proposed OF$^{dist}$ registration performed the best, as is demonstrated visually and by measuring the increase of mutual information in the object of interest and its neighborhood.

Citation: Kateřina Škardová, Tomáš Oberhuber, Jaroslav Tintěra, Radomír Chabiniok. Signed-distance function based non-rigid registration of image series with varying image intensity. Discrete & Continuous Dynamical Systems - S, doi: 10.3934/dcdss.2020386
##### References:

show all references

##### References:
Diagram of the proposed method
Volume elements and the central points
Source image (A), target image (B), the absolute value of their difference before registration (C) and after registration by OF$^{dist}$, MIM and MIM$^{dist}$ (D-F). Parameters of OF$^{dist}$: $N_1 = 200, N_2 = 200, \alpha = 1.0, \beta = 3.25, \gamma = 2.5$
Source image (A), target image (B) and the absolute value of their difference before (C) and after registration by OF, MIM and MIM$^{dist}$ (D-F). Parameters of OF$^{dist}$: $N_1 = 200, N_2 = 200, \alpha = 1.5, \beta = 3.75, \gamma = 3.25$
Absolute value of the difference between the target and source signed-distance function before and after registration by OF$^{dist}$ and MIM$^{dist}$. The signed-distance functions are computed on 10-pixel-wide neighborhood of the edges of the object and set to constant outside the neighborhood
The results of OF$^{dist}$ registration of object marked by green line in 6a, and global MIM registration of the whole scene
Images from the MOLLI sequence with segmented myocardium. Parameters for the outer edge detection: $K_{S_1} = 1.3\cdot 10^{-6}$, $K_{S_2} = 1.3\cdot 10^{-6}$, $K_{T} = 2.3\cdot 10^{-6}$. Parameters for the inner edge detection: $K_{S_1} = K_{S_2} = K_{T} = 9.0\cdot 10^{-6}$
Results of registration of $S_1,S_2$ with target image $T$. Parameters of OF$^{dist}$: $N_1 = 256, N_2 = 218, \alpha = 1.25, \beta = 3.5, \gamma = 3.0$
Norms of difference between the target 3b image and source image 3a before and after registration by OF$^{dist}$, MIM and MIM$^{dist}$
 $\| T - S \|_2$ $\| T - S_{OF^{dist}} \|_2$ $\| T - S_{MIM} \|_2$ $\| T - S_{MIM^{dist}} \|_2$ 9462.47 3130.83 2173.93 2607.66
 $\| T - S \|_2$ $\| T - S_{OF^{dist}} \|_2$ $\| T - S_{MIM} \|_2$ $\| T - S_{MIM^{dist}} \|_2$ 9462.47 3130.83 2173.93 2607.66
Norms of difference between the target image in Figure 4b image and source image in Figure 4a before and after registration by OF$^{dist}$, MIM and MIM$^{dist}$
 $\| T - S \|_2$ $\| T - S_{OF^{dist}} \|_2$ $\| T - S_{MIM} \|_2$ $\| T - S_{MIM^{dist}} \|_2$ 6654.44 2199.07 4121.95 3602.27
 $\| T - S \|_2$ $\| T - S_{OF^{dist}} \|_2$ $\| T - S_{MIM} \|_2$ $\| T - S_{MIM^{dist}} \|_2$ 6654.44 2199.07 4121.95 3602.27
Norms of difference between target and source signed-distance function before and after registration by OF$^{dist}$ and MIM$^{dist}$. The source and target objects can be seen in Figure 4a and 4b, respectively
 $\| \phi_T - \phi_S \|_2$ $\| \phi_T - \phi_{S,{OF^{dist}}} \|_2$ $\| \phi_T - \phi_{S,{MIM^{dist}}} \|_2$ 0.849915 0.239367 0.2950385
 $\| \phi_T - \phi_S \|_2$ $\| \phi_T - \phi_{S,{OF^{dist}}} \|_2$ $\| \phi_T - \phi_{S,{MIM^{dist}}} \|_2$ 0.849915 0.239367 0.2950385
Norms of difference between the target image 4b and source image 4a before registration, after registration of one object by OF$^{dist}$, and after global registration by MIM
 $\| T - S \|_2$ $\| T - S_{1,OF^{dist}} \|_2$ $\| T - S_{MIM} \|_2$ 6654.44 5062.1 4121.95
 $\| T - S \|_2$ $\| T - S_{1,OF^{dist}} \|_2$ $\| T - S_{MIM} \|_2$ 6654.44 5062.1 4121.95
MI target image and source images from Figure 8 before and after registration by OF$^{dist}$ and MIM and MIM$^{dist}$. The MI was computed only in the surroundings of the segmented objects
 $i$ $MI(T,S_i)$ $MI(T,S_{i,OF^{dist}})$ $MI(T,S_{i,MIM})$ $MI(T,S_{i,MIM^{dist}})$ 1 1.1556 1.2184 1.2273 1.2170 2 1.1012 1.2034 1.2052 1.1964
 $i$ $MI(T,S_i)$ $MI(T,S_{i,OF^{dist}})$ $MI(T,S_{i,MIM})$ $MI(T,S_{i,MIM^{dist}})$ 1 1.1556 1.2184 1.2273 1.2170 2 1.1012 1.2034 1.2052 1.1964
 [1] Mehdi Bastani, Davod Khojasteh Salkuyeh. On the GSOR iteration method for image restoration. Numerical Algebra, Control & Optimization, 2021, 11 (1) : 27-43. doi: 10.3934/naco.2020013 [2] Abdelghafour Atlas, Mostafa Bendahmane, Fahd Karami, Driss Meskine, Omar Oubbih. A nonlinear fractional reaction-diffusion system applied to image denoising and decomposition. Discrete & Continuous Dynamical Systems - B, 2020  doi: 10.3934/dcdsb.2020321 [3] Liping Tang, Ying Gao. Some properties of nonconvex oriented distance function and applications to vector optimization problems. Journal of Industrial & Management Optimization, 2021, 17 (1) : 485-500. doi: 10.3934/jimo.2020117 [4] Jia Cai, Guanglong Xu, Zhensheng Hu. Sketch-based image retrieval via CAT loss with elastic net regularization. Mathematical Foundations of Computing, 2020, 3 (4) : 219-227. doi: 10.3934/mfc.2020013 [5] Yifan Chen, Thomas Y. Hou. Function approximation via the subsampled Poincaré inequality. Discrete & Continuous Dynamical Systems - A, 2021, 41 (1) : 169-199. doi: 10.3934/dcds.2020296 [6] Russell Ricks. The unique measure of maximal entropy for a compact rank one locally CAT(0) space. Discrete & Continuous Dynamical Systems - A, 2021, 41 (2) : 507-523. doi: 10.3934/dcds.2020266 [7] Cheng Peng, Zhaohui Tang, Weihua Gui, Qing Chen, Jing He. A bidirectional weighted boundary distance algorithm for time series similarity computation based on optimized sliding window size. Journal of Industrial & Management Optimization, 2021, 17 (1) : 205-220. doi: 10.3934/jimo.2019107 [8] Bahaaeldin Abdalla, Thabet Abdeljawad. Oscillation criteria for kernel function dependent fractional dynamic equations. Discrete & Continuous Dynamical Systems - S, 2020  doi: 10.3934/dcdss.2020443 [9] Lingfeng Li, Shousheng Luo, Xue-Cheng Tai, Jiang Yang. A new variational approach based on level-set function for convex hull problem with outliers. Inverse Problems & Imaging, , () : -. doi: 10.3934/ipi.2020070 [10] Mohammed Abdulrazaq Kahya, Suhaib Abduljabbar Altamir, Zakariya Yahya Algamal. Improving whale optimization algorithm for feature selection with a time-varying transfer function. Numerical Algebra, Control & Optimization, 2021, 11 (1) : 87-98. doi: 10.3934/naco.2020017

2019 Impact Factor: 1.233