Journal of X-Ray Science and Technology - Volume 25, issue 1

Authors: Włodarczyk, Bartłomiej

Article Type: Research Article

Abstract: This article considers the problem of recovering edge-illumination x-ray phase contrast (EIXPC) images from a set of potentially Poisson noisy projection measurements. The authors cast a recovery as a sparse regularization problem based on Anscombe multiscale variance stabilizing transform (MS-VST) with fast discrete curvelet transform which was applied to simulated edge-illumination x-ray phase contrast images. For accurate modelling, the noise characteristics of the EIXPCi data are used to determine the relative importance of each projection. Two implementations of curvelet sparse regularization transforms were applied, including the unequally-spaced fast Fourier transform and the wrapping-based transform. The algorithms were evaluated in terms …of contrast improvement, quality of image restoration, object perceptibility, and peak signal-to-noise ratio. The methods provide nearly optimal solution without excessive memory and recovery time requirement. The performance of the proposed algorithms is demonstrated through a series of complex numerical geometric and anthropomorphic phantom studies. The results of numerical simulations demonstrate that the discrete curvelet transform with MS-VST is fast and robust, and it can effectively improve image quality, preserve and enhance edges and restore lost information while significantly reducing the noise. Additionally, both sparse sampling and decreasing x-ray tube current (i.e. noisy data) lead to the reduction of radiation dose in the x-ray imaging. Show more

Keywords: biomedical imaging, curvelet regularization, x-ray phase-contrast, x-ray imaging

DOI: 10.3233/XST-16190

Citation: Journal of X-Ray Science and Technology, vol. 25, no. 1, pp. 145-170, 2017

Authors: Wang, Yunzhi | Aghaei, Faranak | Zarafshani, Ali | Qiu, Yuchen | Qian, Wei | Zheng, Bin

Article Type: Research Article

Abstract: PURPOSE: To develop a new computer-aided diagnosis (CAD) scheme that computes visually sensitive image features routinely used by radiologists to develop a machine learning classifier and distinguish between the malignant and benign breast masses detected from digital mammograms. METHODS: An image dataset including 301 breast masses was retrospectively selected. From each segmented mass region, we computed image features that mimic five categories of visually sensitive features routinely used by radiologists in reading mammograms. We then selected five optimal features in the five feature categories and applied logistic regression models for classification. A new CAD interface was also …designed to show lesion segmentation, computed feature values and classification score. RESULTS: Areas under ROC curves (AUC) were 0.786±0.026 and 0.758±0.027 when to classify mass regions depicting on two view images, respectively. By fusing classification scores computed from two regions, AUC increased to 0.806±0.025. CONCLUSION: This study demonstrated a new approach to develop CAD scheme based on 5 visually sensitive image features. Combining with a “visual aid” interface, CAD results may be much more easily explainable to the observers and increase their confidence to consider CAD generated classification results than using other conventional CAD approaches, which involve many complicated and visually insensitive texture features. Show more

Keywords: Computer-aided diagnosis (CAD), classification of mammographic masses, quantification of visually sensitive image features, quantitative image feature selection in CAD

DOI: 10.3233/XST-16212

Citation: Journal of X-Ray Science and Technology, vol. 25, no. 1, pp. 171-186, 2017