Segmenting Ultrasound Images of the Prostate Using Neural Networks

Abstract

This paper describes a method for segmenting transrectal ultrasound images of the prostate using feedforward neural networks. Segmenting two-dimensional images of the prostate into prostate and nonprostate regions is required when forming a three-dimensional image of the prostate from a set of parallel two-dimensional images. Three neural network architectures are presented as examples and discussed. Each of these networks was trained using a small portion of a training image segmented by an expert sonographer. The results of applying the trained networks to the entire training image and to adjacent images in the two-dimensional image set are presented and discussed. The final network architecture was also trained with additional data from two other images in the set. The results of applying this retrained network to each of the images in the set are presented and discussed.

A survey of prostate modeling for image analysis

“…Transformation-based approaches, such as the Gabor filter response [66][67][68][69] and wavelet [70][71][72] methods, represent an image in a space in which the coordinate system can be interpreted in a manner that is related to the characteristics of the texture, such as the frequency [73].…”

A reinforcement agent for object segmentation in ultrasound images

“…Due to a very low signal-to-noise ratio, traditional edge detectors fail to extract the correct boundaries (Insana & Brown, 1993;Nanayakkara, Samarabandu, & Fenster, 2006). Consequently, many methods have been introduced to facilitate more accurate automatic or semiautomatic segmentation of the prostate boundaries in these images (Betrounia, Vermandela, Pasquierc, Maoucheb, & Rousseaua, 2005;Chiu, Freeman, Salama, & Fenster, 2004;Gong, Pathak, Haynor, Cho, & Kim, 2004;Ladak et al, 2000;Pathak, Chalana, Haynor, & Kim, 2000;Prater & Richard, 1992;Noble & Boukerroui, 2006;Shen, Zhan, & Davatzikos, 2003;Wang, Cardinal, Downey, & Fenster, 2003). They can usually generate better results by taking into account prior knowledge.…”

Identifying ambiguous prostate gland contours from histology using capsule shape information and least squares curve fitting

“…Despite the numerous research studies in segmenting structures from medical images [1,12,18,21,22] and reconstructing a compact geometric representation of these structures, no study, to the best of our knowledge, has been done to automatically identify the complete prostate capsule in medical images.…”

“…Not only does this assessment allow surgeons to compare the quality of one surgical approach versus another, but also provides an evaluation of surgeons' surgical performances as related to a standard one [14]. Recent studies are focused more on a statistical modelbased segmentation algorithms [2,3,5,13,21] than deformable models [8][9][10][11]17].…”