A Medical Texture Local Binary Pattern For TRUS Prostate Segmentation

Abstract

Prostate cancer diagnosis and treatment rely on segmentation of Transrectal Ultrasound (TRUS) prostate images. This is a challenging and difficult task dut to weak prostate boundaries, speckle noise and the short range of gray levels. Advances in digital imaging techniques have made it possible the acquisition of large volumes of TRUS prostate images so that there is considerable demand for automated segmentation systems. Local Binary Pattern (LBP) has been used for texture segmentation and analysis. Despite its promising performance for texture classification it has not yet been considered for TRUS prostate segmentation. In this paper we introduce a medical texture local binary pattern operator designed for applications of medical imaging where different tissues or micro organisms might maintain extremely weak underlying textures that make it impossible or very difficult for ordinary texture analysis approaches to classify them. In the proposed method the deformations of a level set contour are controlled based on the medical texture local binary pattern operator.

Texture Analysis Methods for Medical Image Characterisation

“…Texture analysis has also been used to identify unique pathology on multi-modality images of cancer patients. Using the local binary operator to analyse the weak underlying textures found in transrectal ultrasound images of the prostate, Kachouie and Fieguth demonstrated that the approach was suitable for segmentation of the prostate (Kachouie & Fieguth, 2007). In another cancer-related study of 48 normal images and 58 cancer images of the colon, Esgiar et al, demonstrated that by adding a fractal feature to traditional statistical features the sensitivity of the classification improved (Esgiar et al, 2002).…”

A survey of prostate modeling for image analysis

“…Rafiee et al [50] Appearance Statistical Mohamed et al [68] Appearance Statistical Song et al [75] Shape and appearance Graph-theoretic framework Liu et al [36] Shape Deformable ellipse Segmentation and clustering Kachouie and Fieguth [63] and Kachouie et al [35] Appearance A medical texture local binary pattern operator Richard et al [60] Appearance Law's mask Classification…”

“…Segmentation is applied to extract the prostate [47,50,63] and to describe its zonal anatomy to permit data reduction and the application of different analysis algorithms based on the considered region.…”

“…In the model-based approach, some parameters (such as the fractal dimension [62], the medical texture local binary pattern (MTLBP) feature [63], and wavelet and spectral features extracted from the radiofrequency time series [64]) are used to characterize the local texture of a region. Zaim [65] conquered neural networks using a feature-based self-organizing map (SOM) formed from the spatial information, gray-level and texture information.…”

Characterisation of radiotherapy planning volumes using textural analysis

“…In a study using the local binary pattern (LBP) operator for texture classification, Kachouie and Fieguth [12] demonstrated that the approach was sensitive to the weak underlying textures present in transrectal ultrasound (TRUS) prostate images. Furthermore, the authors demonstrated that the approach had the necessary discriminative attributes for segmentation of the prostate.…”