Brain MR Image Enhancement for Tumor Segmentation Using 3D U-Net

MRI images are visually inspected by domain experts for the analysis and quantification of the tumorous tissues. Due to the large volumetric data, manual reporting on the images is subjective, cumbersome, and error prone. To address these problems, automatic image analysis tools are employed for tumor segmentation and other subsequent statistical analysis. However, prior to the tumor analysis and quantification, an important challenge lies in the pre-processing. In the present study, permutations of different pre-processing methods are comprehensively investigated. In particular, the study focused on Gibbs ringing artifact removal, bias field correction, intensity normalization, and adaptive histogram equalization (AHE). The pre-processed MRI data is then passed onto 3D U-Net for automatic segmentation of brain tumors. The segmentation results demonstrated the best performance with the combination of two techniques, i.e., Gibbs ringing artifact removal and bias-field correction. The proposed technique achieved mean dice score metrics of 0.91, 0.86, and 0.70 for the whole tumor, tumor core, and enhancing tumor, respectively. The testing mean dice scores achieved by the system are 0.90, 0.83, and 0.71 for the whole tumor, core tumor, and enhancing tumor, respectively. The novelty of this work concerns a robust pre-processing sequence for improving the segmentation accuracy of MR images. The proposed method overcame the testing dice scores of the state-of-the-art methods. The results are benchmarked with the existing techniques used in the Brain Tumor Segmentation Challenge (BraTS) 2018 challenge.

Artifact Reduction in Compressed Sensing Averaging Techniques for High-Resolution Magnetic Resonance Images

This study aims to introduce a new compressed sensing averaging (CSA) technique for the reduction of blurring and/or ringing artifacts, depending on the k-space sampling ratio. A full k-space dataset and three randomly undersampled datasets were obtained for CSA images in a brain phantom and a healthy subject. An additional simulation was performed to assess the effect of the undersampling ratio on the images and the signal-to-noise ratios (SNRs). The image sharpness, spatial resolution, and contrast between tissues were analyzed and compared with other CSA techniques. Compared to CSA with multiple acquisition (CSAM) at 25%, 35%, and 45% undersampling, the reduction rates of the k-space lines of CSA with keyhole (CSAK) were 10%, 15%, and 22%, respectively, and the acquisition time was reduced by 16%, 23%, and 32%, respectively. In the simulation performed with a full sampling k-space dataset, the SNR decreased to 10.41, 9.80, and 8.86 in the white matter and 9.69, 9.35, and 8.46 in the gray matter, respectively. In addition, the ringing artifacts became substantially more predominant as the number of sampling lines decreased. The 50% modulation transfer functions were 0.38, 0.43, and 0.54 line pairs per millimeter for CSAM, CSAK with high-frequency sharing (CSAKS), and CSAK with high-frequency copying (CSAKC), respectively. In this study, we demonstrated that the smaller the sampling line, the more severe the ringing artifact, and that the CSAKC technique proposed to overcome the artifacts that occur when using CSA techniques did not generate artifacts, while it increased spatiotemporal resolution.

DCT Domain Detail Image Enhancement for More Resolved Images

This paper develops a detail image signal enhancement that makes images perceived as being clearer and more resolved and so more effective for higher resolution displays. We observe that the local variant signal enhancement makes images more vivid, and the more revealed granular signals harmonically embedded on the local variant signals make images more resolved. Based on this observation, we develop a method that not only emphasizes the local variant signals by scaling up the frequency energy in accordance with human visual perception, but also strengthens the granular signals by embedding the alpha-rooting enhanced frequency components. The proposed energy scaling method emphasizes the detail signals in texture images and rarely boosts noisy signals in plain images. In addition, to avoid the local ringing artifact, the proposed method adjusts the enhancement direction to be parallel to the underlying image signal direction. It was verified through subjective and objective quality evaluations that the developed method makes images perceived as clearer and highly resolved.

DCT-Domain Detail Image Enhancement for More Resolved Images

This paper develops a detail image signal enhancement that makes images perceived as clearer and more resolved and so is more effective for higher resolution displays. We observe that the local variant signal enhancement makes images more vivid, and the more revealed granular signals harmonically embedded on the local variant signals make images more resolved. Based on this observation, we develop a method that not only emphasizes the local variant signals by scaling up the frequency energy in accordance with human visual perception, but also strengths up the granular signals by embedding the alpha-rooting enhanced frequency components. The proposed energy scaling method emphasizes the detail signals in texture images and rarely boosts noisy signals in plain images. In addition, to avoid the local ringing artifact, the proposed method adjusts the enhancement direction to be parallel to the underlying image signal direction. It was verified through the subjective and objective quality evaluations that the developed method makes images perceived as clearer and highly resolved.

Attention-based Convolutional Neural Network for MRI Gibbs-ringing Artifact Suppression

Gibbs-ringing artifact is a common artifact in MRI image processing. As MRI raw data is taken in a frequency domain, 2D in- verse discrete Fourier transform is applied to visualize data. Inability to take inverse Fourier transform of full spectrum (full k-space) leads to the insufficient sampling of the high frequency data and results in a well-known Gibbs phenomenon. It is worth to notice that truncation of high frequency information generates a significant blur, thus some techniques from other image restoration problems (for example, image deblur task) can be successfully used. We propose attention-based convolutional neural network for Gibbs-ringing reduction which is the extension of recently proposed GAS-CNN (Gibbs-ringing Artifact Suppression Convolutional Neural Network). Proposed method includes simplified non-linear mapping, amended by LRNN (Layer Recurrent Neural Network) refinement block with feature attention module, controlling the correlation between input and output tensors of the refinement unit. The research shows that the proposed post-processing refinement construction considerably simplifies the non-linear mapping.