Early detection of glaucoma is essential to minimizing the risk of visual loss. It has been shown that a good
predictor of glaucoma is the cup-to-disc ratio of the optic nerve head. This paper presents an automated method
to segment the optic disc. Our approach utilizes pixel feature selection to train a feature set to recognize the
region of the disc. Soft pixel classification is used to generate a probability map of the disc. A new cost function
is developed for maximizing the probability of the region within the disc. The segmentation of the image is done
using a novel graph search algorithm capable of detecting the border maximizing the probability of the disc.
The combination of graph search and pixel classification enables us to incorporate large feature sets into the cost
function design, which is critical for segmentation of the optic disc. Our results are validated against a reference
standard of 82 datasets and compared to the manual segmentations of 3 glaucoma fellows.
Early detection of glaucoma is essential to minimizing the risk of visual loss. It has been shown that a good predictor of glaucoma is the cup-to-disc ratio of the optic nerve head. This paper presents a highly automated method to segment the 'rim' (disc) and 'cup' from the optic nerve head in stereo images and calculate the cup-to-disc ratio. In this approach, the optic nerve head is unwrapped in polar coordinates and represented as a graph. Utilizing a novel and efficient graph searching technique for determining globally optimal closed-paths and an intelligent cost function, the rim and the cup are segmented from the stereo images. The results offer a more intuitive quantitative analysis compared to current planimetry-based techniques because the ophthalmologist can view the segmented images along with the derived cup-to-disc ratio.
The accurate segmentation of the brain from three-dimensional medical imagery is important as the basis for visualization, morphometry, surgical planning and intraoperative navigation. The complex and variable nature of brain anatomy makes recognition of the brain boundaries a difficult problem and frustrates segmentation schemes based solely on local image features. We have developed a deformable surface model of the brain as a mechanism for utilizing a priori anatomical knowledge in the segmentation process. The active surface template uses an energy minimization scheme to find a globally consistent surface configuration given a set of potentially ambiguous image features. Solution of the entire 3D problem at once produces superior results to those achieved using a slice by slice approach. We have achieved good results with MR image volumes of both normal and abnormal subjects. Evaluation of the segmentation results has been performed using cadaver studies.
We developed techniques to unobtrusively track direction and pupil diameter of radiologists reading a wide variety of films. This study concentrated on mammography since the use of mammograms is important for successful treatment of breast cancer, and we wished to determine how previous studies of eye gaze with lung films relate to the specialized field of mammography. Our objective was to identify eye gaze patterns in mammographic experts as they observe features, such as masses and microcalcification clusters: identification of these patterns could lead to improving the rate of early detection of breast cancer. Our near IR light system successfully tracked eye gaze direction and pupil diameter of mammographic experts evaluating films. The association of long eye gaze dwells with diagnostic accuracy varied with the type of object being viewed. In films with masses, false positive diagnoses were associated with long dwells: this is similar to published results of observing lung nodule diagnosis. In mammograms with microcalcifications, true positive diagnoses were associated with long dwells. The association of prolonged dwells with true positive diagnoses of microcalcifications is a new observation.
We are developing a method to automatically classify multispectral medical images using context dependent methods. The model is built with the knowledge that cluster of tissue features will overlap in feature space. The goal is to reduce the classification error that results from this cluster overlap. Initialization of the probability of a pixel belonging to a tissue class can take advantage of a priori class distributions if such knowledge exists. Otherwise, the procedure can resort to modeling each class with a Gaussian distribution. These probabilities can then be iteratively updated using either a relaxation labeling algorithm or a Markov random fields algorithm. Once the model converges, iterations cease and each pixel is classified using the maximum probability for all classes.
Traditional, bottom-up segmentation approaches have proven inadequate when faced with the anatomical complexity and variability exhibited by biological structures such as the brain. A 3- D extension to the 'snakes' algorithm has been implemented and used to segment the skin and brain surfaces from MRI image volumes of the head in an effort to investigate model-based, top-down segmentation strategies. These active surfaces allow closed surfaces of complex objects to be recovered using a prior knowledge in the form of initial conditions and applied external 'forces'. Preliminary results suggest that active surfaces may be initialized according to a preconceived model and adaptively deformed by image data to recover the desired object surface.
With the advent of fast 3D magnetic resonance imaging (MRI) sequences, truly 3D volumes of data can be routinely acquired. While images produced by modalities such as positron emission tomography (PET;) and single photon emission computed tomography (SPECT) depict functional information, newer MRI techniques like magnetization-prepared rapid gradient-echo (MP-RAGE) capture a great deal of ana tomical detail [3,10]. Such 3D images are used by clinicians in two types of tasks, visualization and quLantifi cation. Visualization, in its most basic form, permits a user to see structures of interest within a volume of data. The "structures of interest" may not correspond to any physically visible phenomenon (e.g. the quan tity of blood flow to neural tissue in a functional image) but the process of visualization transforms the data into images which may then be displayed using computer graphics. On the other hand, quantification, while often depicted graphically, attempts to reduce the mass of data into numbers useful as clinical indicators. A major obstacle to both of these tasks is the prerequisite image segmentation. In order for the brain to be visualized beneath the overlying head, a segmentation step must determine which volume elements, or voxels, in the 3D head image correspond to the brain. Similarly, quantification requires the distinguishing of background voxels from voxels corresponding to the VOl. A growing number of clinical studies depend on volume measurements after segmentation. Examples include: tracking the progression/remission of disease processes (e.g. the size of intracranial tumors); evaluating the neuroanatomical abnormalities associated with schizophrenia (e.g. ventricular volumes); and determining atrophy associated with Alzheimer-type dementia and temporal lobe epilepsy (as in the hippocampal formation).
Magnetic resonance imaging (MRI) provides excellent soft tissue contrast enabling the non-invasive visualization of soft lissue diseases. The quantification of tissues distinguishable in MR images significantly increases the diagnostic information available to physicians. New 3-D display workstations are available that can also make use of the tissue characteristics to generate clinically useful views of a patient. While simple tissue selection methods work with computed tomography (CT) images these same methods usually do not work with MR images. Several feasibility studies of tissue classification methods have been performed on MR images but few comparative studies of these methods have been published and little work is available on the best statistical model of tissues in MIRI. We have developed a novel method for the identification and quantification of soft tissues from MRI atherosclerosis in particular. This project is part of our work on the development of tissue characterization and identification tools to facilitate soft tissue disease diagnosis and evaluation utilizing MR imagery. Several supervised pattern recognition methods were investigated for tissue identification in MR images such as a Fisher linear discriminant and a minimum distance to the means classifier. For tissue in vivo adequate histology can be difficult to collect. We used cluster analysis methods to generate the necessary training information. ISODATA was modified to use hierarchical stopping rules to determine the true number of tissues in the images. This new method was
ICOS is a modular software system for the 3-dimensional reconstruction of icosohedral particles from 2-dimensional projections. Noise in the resulting reconstructions is due to many sources including ringing artifact due to finite sampling of Fourier transforms and is comparable with many of the desired components representing particulate structure. This paper describes a method for reducing ringing artifact without significant loss of resolution. Application of a Blackman window to the original Fourier transforms significantly improves the signal to noise ratio by suppressing ringing artifact but reduces resolution in the fmal reconstruction. Taking a voxel by voxel minimum of the filtered and unfiltered 3-dimensional reconstruction restores the original resolution while retaining the enhanced SNR. 2.