Real-time image denoising algorithm in teleradiology systems - art.no.606304

Denoising of medical images in wavelet domain has potential application in transmission technologies such as teleradiology. This technique becomes all the more attractive when we consider the progressive transmission in a teleradiology system. The transmitted images are corrupted mainly due to noisy channels. In this paper, we present a new real time image denoising scheme based on limited restoration of bit-planes of wavelet coefficients. The proposed scheme exploits the fundamental property of wavelet transform - its ability to analyze the image at different resolution levels and the edge information associated with each sub-band. The desired bit-rate control is achieved by applying the restoration on a limited number of bit-planes subject to the optimal smoothing. The proposed method adapts itself to the preference of the medical expert; a single parameter can be used to balance the preservation of (expert-dependent) relevant details against the degree of noise reduction. The proposed scheme relies on the fact that noise commonly manifests itself as a fine-grained structure in image and wavelet transform allows the restoration strategy to adapt itself according to directional features of edges. The proposed approach shows promising results when compared with unrestored case, in context of error reduction. It also has capability to adapt to situations where noise level in the image varies and with the changing requirements of medical-experts. The applicability of the proposed approach has implications in restoration of medical images in teleradiology systems. The proposed scheme is computationally efficient.

Design of a FIR filter for image restoration using principal component neural networks

The neural network finds its application in many image denoising applications because of its inherent characteristics such as nonlinear mapping and self-adaptiveness. The design of filters largely depends on the a-priori knowledge about the type of noise. Due to this, standard filters are application and image specific. Widely used filtering algorithms reduce noisy artifacts by smoothing. However, this operation normally results in smoothing of the edges as well. On the other hand, sharpening filters enhance the high frequency details making the image non-smooth. An integrated general approach to design a finite impulse response filter based on principal component neural network (PCNN) is proposed in this study for image filtering, optimized in the sense of visual inspection and error metric. This algorithm exploits the inter-pixel correlation by iteratively updating the filter coefficients using PCNN. This algorithm performs optimal smoothing of the noisy image by preserving high and low frequency features. Evaluation results show that the proposed filter is robust under various noise distributions. Further, the number of unknown parameters is very few and most of these parameters are adaptively obtained from the processed image.

Computationally efficient optical tomographic reconstructions through waveletizing the normalized quadratic perturbation equation - art. no.61640R

In this paper, we present a wavelet - based approach to solve the non-linear perturbation equation encountered in optical tomography. A particularly suitable data gathering geometry is used to gather a data set consisting of differential changes in intensity owing to the presence of the inhomogeneous regions. With this scheme, the unknown image, the data, as well as the weight matrix are all represented by wavelet expansions, thus yielding the representation of the original non - linear perturbation equation in the wavelet domain. The advantage in use of the non-linear perturbation equation is that there is no need to recompute the derivatives during the entire reconstruction process. Once the derivatives are computed, they are transformed into the wavelet domain. The purpose of going to the wavelet domain, is that, it has an inherent localization and de-noising property. The use of approximation coefficients, without the detail coefficients, is ideally suited for diffuse optical tomographic reconstructions, as the diffusion equation removes most of the high frequency information and the reconstruction appears low-pass filtered. We demonstrate through numerical simulations, that through solving merely the approximation coefficients one can reconstruct an image which has the same information content as the reconstruction from a non-waveletized procedure. In addition we demonstrate a better noise tolerance and much reduced computation time for reconstructions from this approach.

Elimination of cross-talk in diffuse optical tomographic reconstructions through a weighted update procedure: results of simulation study - art. no. 61640S

Reconstructions in optical tomography involve obtaining the images of absorption and reduced scattering coefficients. The integrated intensity data has greater sensitivity to absorption coefficient variations than scattering coefficient. However, the sensitivity of intensity data to scattering coefficient is not zero. We considered an object with two inhomogeneities (one in absorption and the other in scattering coefficient). The standard iterative reconstruction techniques produced results, which were plagued by cross talk, i.e., the absorption coefficient reconstruction has a false positive corresponding to the location of scattering inhomogeneity, and vice-versa. We present a method to remove cross talk in the reconstruction, by generating a weight matrix and weighting the update vector during the iteration. The weight matrix is created by the following method: we first perform a simple backprojection of the difference between the experimental and corresponding homogeneous intensity data. The built up image has greater weightage towards absorption inhomogeneity than the scattering inhomogeneity and its appropriate inverse is weighted towards the scattering inhomogeneity. These two weight matrices are used as multiplication factors in the update vectors, normalized backprojected image of difference intensity for absorption inhomogeneity and the inverse of the above for the scattering inhomogeneity, during the image reconstruction procedure. We demonstrate through numerical simulations, that cross-talk is fully eliminated through this modified reconstruction procedure.

Composite MR image reconstruction and unaliasing for general trajectories using neural networks

In rapid parallel magnetic resonance imaging, the problem of image reconstruction is challenging. Here, a novel image reconstruction technique for data acquired along any general trajectory in neural network framework, called ``Composite Reconstruction And Unaliasing using Neural Networks'' (CRAUNN), is proposed. CRAUNN is based on the observation that the nature of aliasing remains unchanged whether the undersampled acquisition contains only low frequencies or includes high frequencies too. Here, the transformation needed to reconstruct the alias-free image from the aliased coil images is learnt, using acquisitions consisting of densely sampled low frequencies. Neural networks are made use of as machine learning tools to learn the transformation, in order to obtain the desired alias-free image for actual acquisitions containing sparsely sampled low as well as high frequencies. CRAUNN operates in the image domain and does not require explicit coil sensitivity estimation. It is also independent of the sampling trajectory used, and could be applied to arbitrary trajectories as well. As a pilot trial, the technique is first applied to Cartesian trajectory-sampled data. Experiments performed using radial and spiral trajectories on real and synthetic data, illustrate the performance of the method. The reconstruction errors depend on the acceleration factor as well as the sampling trajectory. It is found that higher acceleration factors can be obtained when radial trajectories are used. Comparisons against existing techniques are presented. CRAUNN has been found to perform on par with the state-of-the-art techniques. Acceleration factors of up to 4, 6 and 4 are achieved in Cartesian, radial and spiral cases, respectively. (C) 2010 Elsevier Inc. All rights reserved.

Accelerating frequency-domain diffuse optical tomographic image reconstruction using graphics processing units

Diffuse optical tomographic image reconstruction uses advanced numerical models that are computationally costly to be implemented in the real time. The graphics processing units (GPUs) offer desktop massive parallelization that can accelerate these computations. An open-source GPU-accelerated linear algebra library package is used to compute the most intensive matrix-matrix calculations and matrix decompositions that are used in solving the system of linear equations. These open-source functions were integrated into the existing frequency-domain diffuse optical image reconstruction algorithms to evaluate the acceleration capability of the GPUs (NVIDIA Tesla C 1060) with increasing reconstruction problem sizes. These studies indicate that single precision computations are sufficient for diffuse optical tomographic image reconstruction. The acceleration per iteration can be up to 40, using GPUs compared to traditional CPUs in case of three-dimensional reconstruction, where the reconstruction problem is more underdetermined, making the GPUs more attractive in the clinical settings. The current limitation of these GPUs in the available onboard memory (4 GB) that restricts the reconstruction of a large set of optical parameters, more than 13, 377. (C) 2010 Society of Photo-Optical Instrumentation Engineers. DOI: 10.1117/1.3506216]

Evaluation of 2D image reconstruction using fan-beam FBP algorithm with no backprojection weight from WLP completed truncated projection data

This paper presents the image reconstruction using the fan-beam filtered backprojection (FBP) algorithm with no backprojection weight from windowed linear prediction (WLP) completed truncated projection data. The image reconstruction from truncated projections aims to reconstruct the object accurately from the available limited projection data. Due to the incomplete projection data, the reconstructed image contains truncation artifacts which extends into the region of interest (ROI) making the reconstructed image unsuitable for further use. Data completion techniques have been shown to be effective in such situations. We use windowed linear prediction technique for projection completion and then use the fan-beam FBP algorithm with no backprojection weight for the 2-D image reconstruction. We evaluate the quality of the reconstructed image using fan-beam FBP algorithm with no backprojection weight after WLP completion.

Fast image reconstruction for fluorescence microscopy

Real-time image reconstruction is essential for improving the temporal resolution of fluorescence microscopy. A number of unavoidable processes such as, optical aberration, noise and scattering degrade image quality, thereby making image reconstruction an ill-posed problem. Maximum likelihood is an attractive technique for data reconstruction especially when the problem is ill-posed. Iterative nature of the maximum likelihood technique eludes real-time imaging. Here we propose and demonstrate a compute unified device architecture (CUDA) based fast computing engine for real-time 3D fluorescence imaging. A maximum performance boost of 210x is reported. Easy availability of powerful computing engines is a boon and may accelerate to realize real-time 3D fluorescence imaging. Copyright 2012 Author(s). This article is distributed under a Creative Commons Attribution 3.0 Unported License. http://dx.doi.org/10.1063/1.4754604]

Benchmarking recognition results on camera captured word image data sets

We have benchmarked the maximum obtainable recognition accuracy on five publicly available standard word image data sets using semi-automated segmentation and a commercial OCR. These images have been cropped from camera captured scene images, born digital images (BDI) and street view images. Using the Matlab based tool developed by us, we have annotated at the pixel level more than 3600 word images from the five data sets. The word images binarized by the tool, as well as by our own midline analysis and propagation of segmentation (MAPS) algorithm are recognized using the trial version of Nuance Omnipage OCR and these two results are compared with the best reported in the literature. The benchmark word recognition rates obtained on ICDAR 2003, Sign evaluation, Street view, Born-digital and ICDAR 2011 data sets are 83.9%, 89.3%, 79.6%, 88.5% and 86.7%, respectively. The results obtained from MAPS binarized word images without the use of any lexicon are 64.5% and 71.7% for ICDAR 2003 and 2011 respectively, and these values are higher than the best reported values in the literature of 61.1% and 41.2%, respectively. MAPS results of 82.8% for BDI 2011 dataset matches the performance of the state of the art method based on power law transform.

Advances in Image Reconstruction Methods for Real-Time Magnetic Resonance Thermometry

Magnetic Resonance Imaging (MRI) has been widely used in cancer treatment planning, which takes the advantage of high-resolution and high-contrast provided by it. The raw data collected in the MRI can also be used to obtain the temperature maps and has been explored for performing MR thermometry. This review article describes the methods that are used in performing MR thermometry, with an emphasis on reconstruction methods that are useful to obtain these temperature maps in real-time for large region of interest. This article also proposes a prior-image constrained reconstruction method for temperature reconstruction in MR thermometry, and a systematic comparison using ex-vivo tissue experiments with state of the art reconstruction method is presented.

Super Resolution Using a Single Image Dictionary

To perform super resolution of low resolution images, state-of-the-art methods are based on learning a pair of lowresolution and high-resolution dictionaries from multiple images. These trained dictionaries are used to replace patches in lowresolution image with appropriate matching patches from the high-resolution dictionary. In this paper we propose using a single common image as dictionary, in conjunction with approximate nearest neighbour fields (ANNF) to perform super resolution (SR). By using a common source image, we are able to bypass the learning phase and also able to reduce the dictionary from a collection of hundreds of images to a single image. By adapting recent developments in ANNF computation, to suit super-resolution, we are able to perform much faster and accurate SR than existing techniques. To establish this claim, we compare the proposed algorithm against various state-of-the-art algorithms, and show that we are able to achieve b etter and faster reconstruction without any training.

B-Spline potential function for maximum a-posteriori image reconstruction in fluorescence microscopy

An iterative image reconstruction technique employing B-Spline potential function in a Bayesian framework is proposed for fluorescence microscopy images. B-splines are piecewise polynomials with smooth transition, compact support and are the shortest polynomial splines. Incorporation of the B-spline potential function in the maximum-a-posteriori reconstruction technique resulted in improved contrast, enhanced resolution and substantial background reduction. The proposed technique is validated on simulated data as well as on the images acquired from fluorescence microscopes (widefield, confocal laser scanning fluorescence and super-resolution 4Pi microscopy). A comparative study of the proposed technique with the state-of-art maximum likelihood (ML) and maximum-a-posteriori (MAP) with quadratic potential function shows its superiority over the others. B-Spline MAP technique can find applications in several imaging modalities of fluorescence microscopy like selective plane illumination microscopy, localization microscopy and STED. (C) 2015 Author(s).

Synthetic image super resolution using FeatureMatch

In this paper, we propose a super resolution (SR) method for synthetic images using FeatureMatch. Existing state-of-the-art super resolution methods are learning based methods, where a pair of low-resolution and high-resolution dictionary pair are trained, and this trained pair is used to replace patches in low-resolution image with appropriate matching patches from the high-resolution dictionary. In this paper, we show that by using Approximate Nearest Neighbour Fields (ANNF), and a common source image, we can by-pass the learning phase, and use a single image for dictionary. Thus, reducing the dictionary from a collection obtained from hundreds of training images, to a single image. We show that by modifying the latest developments in ANNF computation, to suit super resolution, we can perform much faster and more accurate SR than existing techniques. To establish this claim we will compare our algorithm against various state-of-the-art algorithms, and show that we are able to achieve better and faster reconstruction without any training phase.

PolyMage: Automatic Optimization for Image Processing Pipelines

This paper presents the design and implementation of PolyMage, a domain-specific language and compiler for image processing pipelines. An image processing pipeline can be viewed as a graph of interconnected stages which process images successively. Each stage typically performs one of point-wise, stencil, reduction or data-dependent operations on image pixels. Individual stages in a pipeline typically exhibit abundant data parallelism that can be exploited with relative ease. However, the stages also require high memory bandwidth preventing effective utilization of parallelism available on modern architectures. For applications that demand high performance, the traditional options are to use optimized libraries like OpenCV or to optimize manually. While using libraries precludes optimization across library routines, manual optimization accounting for both parallelism and locality is very tedious. The focus of our system, PolyMage, is on automatically generating high-performance implementations of image processing pipelines expressed in a high-level declarative language. Our optimization approach primarily relies on the transformation and code generation capabilities of the polyhedral compiler framework. To the best of our knowledge, this is the first model-driven compiler for image processing pipelines that performs complex fusion, tiling, and storage optimization automatically. Experimental results on a modern multicore system show that the performance achieved by our automatic approach is up to 1.81x better than that achieved through manual tuning in Halide, a state-of-the-art language and compiler for image processing pipelines. For a camera raw image processing pipeline, our performance is comparable to that of a hand-tuned implementation.

FIVDL: Fingerprint Image Verification using Dictionary Learning

Fingerprints are used for identification in forensics and are classified into Manual and Automatic. Automatic fingerprint identification system is classified into Latent and Exemplar. A novel Exemplar technique of Fingerprint Image Verification using Dictionary Learning (FIVDL) is proposed to improve the performance of low quality fingerprints, where Dictionary learning method reduces the time complexity by using block processing instead of pixel processing. The dynamic range of an image is adjusted by using Successive Mean Quantization Transform (SMQT) technique and the frequency domain noise is reduced using spectral frequency Histogram Equalization. Then, an adaptive nonlinear dynamic range adjustment technique is utilized to determine the local spectral features on corresponding fingerprint ridge frequency and orientation. The dictionary is constructed using spatial fundamental frequency that is determined from the spectral features. These dictionaries help in removing the spurious noise present in fingerprints and reduce the time complexity by using block processing instead of pixel processing. Further, dictionaries are used to reconstruct the image for matching. The proposed FIVDL is verified on FVC database sets and Experimental result shows an improvement over the state-of-the-art techniques. (C) 2015 The Authors. Published by Elsevier B.V.