Hong-Ren Su

A method for generating reproducible evidence in fMRI studies.

Insights into cognitive neuroscience from neuroimaging techniques are now required to go beyond the localisation of well-known cognitive functions. Fundamental to this is the notion of reproducibility of experimental outcomes. This paper addresses the central issue that functional magnetic resonance imaging (fMRI) experiments will produce more desirable information if researchers begin to search for reproducible evidence rather than only p value significance. The study proposes a methodology for investigating reproducible evidence without conducting separate fMRI experiments. The reproducible evidence is gathered from the separate runs within the study. The associated empirical Bayes and ROC extensions of the linear model provide parameter estimates to determine reproducibility. Empirical applications of the methodology suggest that reproducible evidence is robust to small sample sizes and sensitive to both the magnitude and persistency of brain activation. It is demonstrated that research findings in fMRI studies would be more compelling with supporting reproducible evidence in addition to standard hypothesis testing evidence.

CT-MR image registration in 3d k-space based on fourier moment matching

CT-MRI registration is a common processing procedure for clinical diagnosis and therapy. We propose a novel K-space affine image registration algorithm via Fourier moment matching. The proposed algorithm is based on estimating the affine matrix from the moment relationship between the corresponding Fourier spectrums. This estimation strategy is very robust because the energy of the Fourier spectrum is mostly concentrated in the low-frequency band, thus the moments of the Fourier spectrum are robust against noises and outliers. Our experiments on the real CT and MRI datasets show that the proposed Fourier-based registration algorithm provides higher registration accuracy than the existing mutual information registration technique.

3D object detection and pose estimation from depth image for robotic bin picking

In this paper, we present a system for automatic object detection and pose estimation from a single depth map containing multiple objects for bin-picking applications. The proposed object detection algorithm is based on matching the keypoints extracted from the depth image by using the RANSAC algorithm with the spin image descriptor. In the proposed system, we combine the keypoint detection and the RANSAC algorithm to detect the objects, followed by the ICP algorithm to refine the 3D pose estimation. In addition, we implement the proposed algorithm on the GPGPU platform to speed-up the computation. Experimental results on simulated depth data are shown to demonstrate the proposed system.

Pedestrian image segmentation via shape-prior constrained random walks

In this paper, we present an automatic and accurate pedestrian segmentation algorithm by incorporating pedestrian shape prior into the random walks segmentation algorithm. The random walks [1] algorithm requires user-specified labels to produce segmentation with each pixel assigned to a label, and it can provide satisfactory segmentation result with proper input labeled seeds. To take advantage of this interactive segmentation algorithm, we improve the random walks segmentation algorithm by incorporating prior shape information into the same optimization formulation. By using the human shape prior, we develop a fully automatic pedestrian image segmentation algorithm. Our experimental results demonstrate that the proposed algorithm significantly outperforms the previous segmentation methods in terms of pedestrian segmentation accuracy on a number of real images.

Non-rigid registration of images with geometric and photometric deformation by using local affine Fourier-moment matching

Registration between images taken with different cameras, from different viewpoints or under different lighting conditions is a challenging problem. It needs to solve not only the geometric registration problem but also the photometric matching problem. In this paper, we propose to estimate the integrated geometric and photometric transformations between two images based on a local affine Fourier-moment matching framework, which is developed to achieve deformable registration. We combine the local Fourier moment constraints with the smoothness constraints to determine the local affine transforms in a hierarchal block model. Our experimental results on registering some real images related by large color and geometric transformations show the proposed registration algorithm provides superior image registration results compared to the state-of-the-art image registration methods.

Beyond p-values: averaged and reproducible evidence in fMRI experiments

In functional magnetic resonance imaging studies, there might exist activation regions routinely involved in experimental sessions, but modest in response magnitude. These regions may not be easily detectable by the conventional p-value approach using a rigid threshold. With particular reference to the reproducibility analysis method proposed in Liou and colleagues, this study presents some within- and between-subject brain-activation patterns that are replicable between experimental modalities, and robust to the method used for generating the patterns. There is a neurophysiological basis behind these reproducible patterns, and the conventional p-value approach using averaged data across subjects might not suggest the complete patterns. For example, recent studies based on the group-averaged data showed a task-induced deactivation in the precuneus and posterior cingulate, but our reproducibility analysis suggests both increased and decreased responses in the two regions. The increased responses localize in these regions with differentially distributed patterns for individual subjects and for different experimental tasks. In this study, we discuss the neurophysiological basis of the reproducible patterns and propose some applications of our research findings to scientific and clinical studies.

Human segmentation from video by combining random walks with human shape prior adaption

In this paper, we propose an automatic human segmentation algorithm for video conferencing applications. Since humans are the principal subject in these videos, the proposed framework is based on human shape clues to separate humans from complex background and replace or blur the background for immersive communication. We first detect face position and size, track human boundary across frames, and propagate the segmentation likelihood to the next frame for obtaining the trimap to be used as input to the Random Walk algorithm. In addition, we also include gradient magnitude in edge weight to enhance the Random Walk segmentation results. Finally, we demonstrate experimental results on several image sequences to show the effectiveness and robustness of the proposed method.

3D Spinal Cord and Nerves Segmentation from STIR-MRI

In this paper, we present a system for spinal cord and nerves segmentation from STIR-MRI. We propose an user interactive segmentation method for 3D images, which is extended from the 2D random walker algorithm and implemented with a slice-section strategy. After obtaining the 3D segmentation result, we build the 3D spinal cord and nerves model for each view using VTK, which is an open-source, freely available software. Then we obtain the point cloud of the spinal cord and nerves surface by registering the three surface models constructed from three STIR-MRI images of different directions. In the experimental results, we show the 3D segmentation results of spinal cord and nerves from the STIR-MRI (Short Tau Inversion Recovery - Magnetic Resonance Imaging)images in three different views, and also display the reconstructed 3D surface model.

A deep learning approach towards pore extraction for high-resolution fingerprint recognition

As high-resolution fingerprint images are becoming more common, the pores have been found to be one of the promising candidates in improving the performance of automated fingerprint identification systems (AFIS). This paper proposes a deep learning approach towards pore extraction. It exploits the feature learning and classification capability of convolutional neural networks (CNNs) to detect pores on fingerprints. Besides, this paper also presents a unique affine Fourier moment-matching (AFMM) method of matching and fusing the scores obtained for three different fingerprint features to deal with both local and global linear distortions. Combining the two aforementioned contributions, an EER of 3.66% can be observed from the experimental results.

Depth image super-resolution via multi-frame registration and deep learning

In this paper, we develop an algorithm for depth image super-resolution from RGB-D images, which are acquired under different imaging conditions so that we can combine them to improve the image quality with precise 3D registration. We focus on how to increase the resolution and quality of depth images by combining multiple RGB-D images and using the deep learning technique. In the proposed solution, we combine multiple RGB-D images by 3D alignment from 3D feature point correspondences and apply the guided filter as the input to SRCNN to obtain the up-sampled depth images. We show depth quality improvement of the up-sampled depth maps by using the proposed algorithm over the traditional methods through experimental results on some public-domain RGB-D datasets.