Showing papers by "Rafael Molina published in 2009"

Variational Bayesian Blind Deconvolution Using a Total Variation Prior

Abstract: In this paper, we present novel algorithms for total variation (TV) based blind deconvolution and parameter estimation utilizing a variational framework. Using a hierarchical Bayesian model, the unknown image, blur, and hyperparameters for the image, blur, and noise priors are estimated simultaneously. A variational inference approach is utilized so that approximations of the posterior distributions of the unknowns are obtained, thus providing a measure of the uncertainty of the estimates. Experimental results demonstrate that the proposed approaches provide higher restoration performance than non-TV-based methods without any assumptions about the unknown hyperparameters.

Bayesian Super-Resolution image reconstruction using an ℓ1 prior

Abstract: This paper deals with the problem of high-resolution (HR) image reconstruction, from a set of degraded, under-sampled, shifted and rotated images, under the Bayesian paradigm, utilizing a variational approximation. Bayesian methods rely on image models that encapsulate prior image knowledge and avoid the ill-posedness of the image restoration problems. In this paper a new prior based on the l1 norm of vertical and horizontal first order differences of image pixel values is introduced and its parameters are estimated. The estimated HR images are compared with images provided by other HR reconstruction methods.

High preoperative serum vascular endothelial growth factor levels predict poor clinical outcome after curative resection of gastric cancer.

Abstract: Background: Tumour vascular endothelial growth factor (VEGF) and tumour urokinase-type plasminogen activator (uPA) are prognostic factors in gastric cancer but surgical specimens are required for testing. The prognostic value of preoperative serum VEGF (s-VEGF) and serum uPA (s-uPA) levels was evaluated in patients undergoing potentially curative (R0) gastric cancer resection. Methods: Concentrations of s-VEGF and s-uPA were measured 97 patients with gastric cancer and 20 controls. Angiogenesis was measured in vitro based on human endothelial cell tube formation. Results: Levels of s-VEGF were higher in patients with gastric cancer than controls (median 288 versus 189 pg/ml respectively; P = 0·002). They were associated with pathological tumour node metastasis (pTNM) stage, pT, pN, lymph node ratio and perineural invasion, and correlated with platelet counts. In multivariable analysis, s-VEGF over 320 pg/ml was the only preoperative predictor of both recurrence and disease-specific survival. Serum from patients with raised s-VEGF levels enhanced angiogenesis in vitro significantly more than serum from those with a s-VEGF level of 320 pg/ml or less. Conclusion: High preoperative s-VEGF level is an independent prognostic factor for recurrence and survival after R0 resection of gastric cancer. This may provide a useful guide to decision making regarding neoadjuvant and adjuvant therapies. Copyright © 2009 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.

Significance of serum CA125 and TPS antigen levels for determination of overall survival after three chemotherapy courses in ovarian cancer patients during long-term follow-up

Abstract: PURPOSE OF INVESTIGATION: To evaluate the prognostic significance for overall survival rate for the marker combination TPS and CA125 in ovarian cancer patients after three chemotherapy courses during long-term clinical follow-up. METHODS: The overall survival of 212 (out of 213) ovarian cancer patients (FIGO Stages I-IV) was analyzed in a prospective multicenter study during a 10-year clinical follow-up by univariate and multivariate analysis. RESULTS: In patients with ovarian cancer FIGO Stage I (34 patients) or FIGO Stage II (30 patients) disease, the univariate and multivariate analysis of the 10-year overall survival data showed that CA125 and TPS serum levels were not independent prognostic factors. In the FIGO Stage III group (112 patients), the 10-year overall survival was 15.2%; while in the FIGO Stage IV group (36 patients) a 10-year overall survival of 5.6% was seen. Here, the tumor markers CA125 and TPS levels were significant prognostic factors in both univariate and multivariate analysis (p < 0.0001). In a combined FIGO Stage III + FIGO Stage IV group (60 patients with optimal debulking surgery), multivariate analysis demonstrated that CA125 and TPS levels were independent prognostic factors. For patients in this combined FIGO Stage III + IV group having both markers below respective discrimination level, 35.3% survived for more than ten years, as opposed to patients having one marker above the discrimination level where the 10-year survival was reduced to 10% of the patients. For patients showing both markers above the respective discrimination level, none of the patients survived for the 10-year follow-up time. CONCLUSION: In FIGO III and IV ovarian cancer patients, only patients with CA 125 and TPS markers below the discrimination level after three chemotherapy courses indicated a favorable prognosis. Patients with an elevated level of CA 125 or TPS or both markers after three chemotherapy courses showed unfavorable prognosis.

Compressive sensing of light fields

Abstract: We propose a novel camera design for light field image acquisition using compressive sensing. By utilizing a randomly coded non-refractive mask in front of the aperture, incoherent measurements of the light passing through different regions are encoded in the captured images. A novel reconstruction algorithm is proposed to recover the original light field image from these acquisitions. Using the principles of compressive sensing, we demonstrate that light field images with high angular dimension can be captured with only a few acquisitions. Moreover, the proposed design provides images with high spatial resolution and signal-to-noise-ratio (SNR), and therefore does not suffer from limitations common to existing light-field camera designs. Experimental results demonstrate the efficiency of the proposed system.

Fast bayesian compressive sensing using Laplace priors

Abstract: In this paper we model the components of the compressive sensing (CS) problem using the Bayesian framework by utilizing a hierarchical form of the Laplace prior to model sparsity of the unknown signal. This signal prior includes some of the existing models as special cases and achieves a high degree of sparsity. We develop a constructive (greedy) algorithm resulting from this formulation where necessary parameters are estimated solely from the observation and therefore no user-intervention is needed. We provide experimental results with synthetic 1D signals and images, and compare with the state-of-the-art CS reconstruction algorithms demonstrating the superior performance of the proposed approach.

Bayesian compressive sensing of wavelet coefficients using multiscale Laplacian priors

Abstract: In this paper, we propose a novel algorithm for image reconstruction from compressive measurements of wavelet coefficients. By incorporating independent Laplace priors on separate wavelet sub-bands, the inhomogeneity of wavelet coefficient distributions and therefore the structural sparsity within images are modeled effectively. We model the problem by adopting a Bayesian formulation, and develop a fast greedy reconstruction algorithm. Experimental results demonstrate that the reconstruction performance of the proposed algorithm is competitive with state-of-the-art methods while outperforming them in terms of running times.

Super-Resolution of Multispectral Images

Abstract: In this paper we propose and analyze a globally and locally adaptive super-resolution Bayesian methodology for pansharpening of multispectral images. The methodology incorporates prior knowledge on the expected characteristics of the multispectral images uses the sensor characteristics to model the observation process of both panchromatic and multispectral images and includes information on the unknown parameters in the model in the form of hyperprior distributions. Using real and synthetic data, the pansharpened multispectral images are compared with the images obtained by other pansharpening methods and their quality is assessed both qualitatively and quantitatively.

Non-convex priors in Bayesian compressed sensing

Abstract: We propose a novel Bayesian formulation for the reconstruction from compressed measurements. We demonstrate that high-sparsity enforcing priors based on l p -norms, with 0 < p ≤ 1, can be used within a Bayesian framework by majorization-minimization methods. By employing a fully Bayesian analysis of the compressed sensing system and a variational Bayesian analysis for inference, the proposed framework provides model parameter estimates along with the unknown signal, as well as the uncertainties of these estimates. We also show that some existing methods can be derived as special cases of the proposed framework. Experimental results demonstrate the high performance of the proposed algorithm in comparison with commonly used methods for compressed sensing recovery.

Super resolution of multispectral images using ℓ1 image models and interband correlations

Abstract: In this paper we propose a novel super-resolution based algorithm for the pansharpening of multispectral images. Within the Bayesian formulation, the proposed methodology incorporates prior knowledge on the expected characteristics of multispectral images; that is, imposes smoothness within each band by means of the energy associated to the l1 norm of vertical and horizontal first order differences of image pixel values and also takes into account the correlation between the bands of the multispectral image. The observation process is modeled using the sensor characteristics of both panchromatic and multispectral images. The method is tested on real and synthetic images, compared with other pan-sharpening methods, and its quality is assessed both qualitatively and quantitatively.

L1 prior majorization in Bayesian image restoration

Abstract: Bayesian methods rely on image priors that encapsulate prior image knowledge and avoid the ill-posedness of image restoration problems. In this paper a new prior based on the l1 norm of vertical and horizontal first order differences of image pixel values is introduced and its parameters are estimated . The results obtained from its application studied and compared with the ones provided by other methods in the literature.

Parameter Estimation in Bayesian Super-Resolution Image Reconstruction from Low Resolution Rotated and Translated Images

Abstract: This paper deals with the problem of high-resolution (HR) image reconstruction, from a set of degraded, under-sampled, shifted and rotated images, utilizing the variational approximation within the Bayesian paradigm. The proposed inference procedure requires the calculation of the covariance matrix of the HR image given the LR observations and the unknown hyperparameters of the probabilistic model. Unfortunately the size and complexity of such matrix renders its calculation impossible, and we propose and compare three alternative approximations. The estimated HR images are compared with images provided by other HR reconstruction methods.

Bayesian blind deconvolution from differently exposed image pairs

Abstract: Photographs acquired under low-light conditions require long exposure times and therefore exhibit significant blurring due to the shaking of the camera. Using shorter exposure times results in sharper images but with a very high level of noise. In this paper we address this problem and present a novel blind deconvolution algorithm for a pair of differently exposed images. We formulate the problem in a hierarchical Bayesian framework by utilizing prior knowledge on the unknown image and blur, and also on the dependency between two observed images. By incorporating a fully Bayesian analysis, the developed algorithm estimates all necessary algorithm parameters along with the unknowns, such that no user-intervention is needed. Moreover, we employ a variational Bayesian inference procedure, which allows for the statistical compensation of errors occurring at different stages of the restoration, and also provides uncertainties of the estimates. Experimental results demonstrate the high restoration performance of the proposed algorithm.

Local Bayesian image restoration using variational methods and Gamma-Normal distributions

Abstract: In this paper we present a new Bayesian methodology for the restoration of blurred and noisy images. Bayesian methods rely on image priors that encapsulate prior image knowledge and avoid the ill-posedness of image restoration problems. We use a spatially varying image prior utilizing a Gamma-Normal hyperprior distribution on the local precision parameters. This kind of hyperprior distribution, which to our knowledge has not been used before in image restoration, allows for the incorporation of information on local as well as global image variability, models correlation of the local precision parameters and is a conjugate hyperprior to the image model used in the paper. The proposed restoration technique is compared with other image restoration approaches, demonstrating its improved performance.

Image restoration by mixture modelling of an overcomplete linear representation

Abstract: We present a new image restoration method based on modelling the coefficients of an overcomplete wavelet response to natural images with a mixture of two Gaussian distributions, having non-zero and zero mean respectively, and reflecting the assumption that this response is close to be sparse. Including the observation model, the resulting procedure iterates between image reconstruction from the hard-thresholding of the response to the current estimate and a fast blur compensation step. Results indicate that our method compares favorably with current wavelet-based restoration methods.

Super resolution of multispectralimages using £1image models and interband correlations

Abstract: In this paper we propose a novel super-resolution based al­ gorithm for the pansharpening of multispectral images. Within the Bayesian formulation, the proposed methodology incorporates prior knowledge on the expected characteristics of multispectral images; that is, imposes smoothness within each band by means of the energy associated to the £1 norm of vertical and horizontal first order differences of image pixel values and also takes into ac­ count the correlation between the bands of the multispectral image. The observation process is modeled using the sensor characteris­ tics of both panchromatic and multispectral images. The method is tested on real and synthetic images, compared with other pan­ sharpening methods, and its quality is assessed both qualitatively and quantitatively.