Dissertation Abstract

A Seed-Based Semi-Automatic Segmentation Method for Serial Section Images

Our understanding of the world around us and the many objects that we encounter is based primarily on three-dimensional information. It is simply part of the environment in which we live and the intuitive nature of our interpretation of our surroundings.

Medical or biological image information is most often collected in the form of two-dimensional, serial section images, such as magnetic resonance imaging (MRI). In such cases, it is difficult for the observer to mentally build and understand the three dimensional structure of the object. Although most image rendering software packages allow for 3D views of the serial sections, they lack the ability to segment, or isolate different objects in the data set. Typically the task of segmentation is performed by highly trained persons who tediously outline the object of interest in each image slice containing the object. It remains a difficult challenge for computers to understand an image and aid in the segmentation process.

Segmentation is the key to creating 3D renderings of distinct objects from serial slice images. This dissertation describes a seed-based semi-automatic segmentation method for objects recorded with serial section images. The simple and robust algorithm is initiated by the creation of a user-defined object seed from a single image slice in the data set. Important seed parameters are then used to automatically segment the object in the remaining images of the data set. These parameters set the contrast enhancement, threshold values, and objects labels that guide the automatic segmentation process.

Mathematically defined shapes are used to verify the performance of the segmentation method. The algorithm is then applied to various biological samples, including the visual neurons of the housefly, Musca domestica . It is the study of this insect's visual system that provides the opportunity to develop this segmentation algorithm. It is hoped that the 3D renderings of the fly visual neurons will lead to improved machine vision systems.

In general, the described segmentation method can be applied to any high contrast serial slice data set that is well aligned and registered. The medical field alone has many applications for rapid generation of 3D segmented models from MRI and other medical imaging modalities.

This dissertation citation and abstract are published with permission of ProQuest Information and Learning. Further reproduction is prohibited without permission.