Colm Boylan

A prospective study of tumor and technical factors associated with positive margins in breast-conservation therapy for nonpalpable malignancy.

BACKGROUND The purpose of this study was to identify factors that predict an increased risk of a positive surgical margin after breast-conserving therapy for nonpalpable carcinoma of the breast. METHODS In this prospective study, 305 patients with nonpalpable invasive breast cancer or ductal carcinoma in situ were identified and underwent localization lumpectomy. Patient, technical, and tumor factors with a potential to predict margin status were documented. RESULTS A 20% positive margin rate was observed. Univariate analysis of patient, tumor, and technical factors revealed that localizations performed under stereotactic guidance (P < .001), presence of in situ disease, high tumor grade, larger tumor size, multifocal disease, and presence of mammographic microcalcifications (P < .02) were predictive of positive margins. With the exception of tumor grade and mammographic microcalcifications, multivariable analysis identified the same factors. CONCLUSIONS This study identified several factors associated with positive margins that should be considered when planning breast-conserving therapy for nonpalpable tumors.

Optimizing the Level Set Algorithm for Detecting Object Edges in MR and CT Images

Specifying the boundary of tissues or an organ is one of the most frequently required tasks for a radiologist. It is a first step for further processing, such as comparing two serial images in time, volume measurements. In the present work, we use genetic algorithms (GA) and where necessary apply a ldquodynamic genetic algorithmrdquo (dGA) procedure, which (we believe) is a unique application, to assess different values for finding an optimal set of parameters that characterize the level set method, a geometric active contour, for use as a boundary detection method. Four quantitative measures are used in calculating geometric differences between the object boundaries, as determined by the level set method, and the desired object boundaries. A semi-automated method is also developed to find the desired boundary for the object. A two-step method requires the user to manipulate the object boundaries obtained by applying an edge detection method based on the Canny filter. By setting the level set method parameters using the output of the GA we obtain accurate boundaries of organs automatically and rapidly.

Review of Texture Quantification of CT Images for Classification of Lung Diseases

Computer-based identification of abnormal regions and classification of diseases using CT images of the lung has been a goal of many investigators. In this paper, we review research that has used texture analysis along with segmentation and fractal analysis. First, a review of texture methods is performed. Recent research on quantitative analysis of the lung using texture methods is categorized into six groups of computational methods: structural, statistical, model based, transform domain, texture-segmentation, and texture-fractal analysis. Finally, the applications of texture-based methods combined with either segmentation algorithms or fractal analysis is evaluated on lung CT images from patients with diseases such as emphysema, COPD, and cancer. We also discuss applications of artificial neural networks, support vector machine, k-nearest, and Bayesian methods to classify normal and diseased segments of CT images of the lung. A combination of these texture methods followed by classifiers could lead to efficient and accurate diagnosis of pulmonary diseases such as pulmonary fibrosis, emphysema, and cancer.