Showing papers by "Paul Sajda published in 2002"

Learning contextual relationships in mammograms using a hierarchical pyramid neural network

Abstract: This paper describes a pattern recognition architecture, which we term hierarchical pyramid/neural network (HPNN), that learns to exploit image structure at multiple resolutions for detecting clinically significant features in digital/digitized mammograms. The HPNN architecture consists of a hierarchy of neural networks, each network receiving feature inputs at a given scale as well as features constructed by networks lower in the hierarchy. Networks are trained using a novel error function for the supervised learning of image search/detection tasks when the position of the objects to be found is uncertain or ill defined. We have evaluated the HPNN's ability to eliminate false positive (FP) regions of interest generated by the University of Chicago's Computer-aided diagnosis (CAD) systems for microcalcification and mass detection. Results show that the HPNN architecture, trained using the uncertain object position (UOP) error function, reduces the FP rate of a mammographic CAD system by approximately 50% without significant loss in sensitivity. Investigation into the types of FPs that the HPNN eliminates suggests that the pattern recognizer is automatically learning and exploiting contextual information. Clinical utility is demonstrated through the evaluation of an integrated system in a clinical reader study. We conclude that the HPNN architecture learns contextual relationships between features at multiple scales and integrates these features for detecting microcalcifications; and breast masses.

Multi-resolution and wavelet representations for identifying signatures of disease.

Abstract: Identifying physiological and anatomical signatures of disease in signals and images is one of the fundamental challenges in biomedical engineering. The challenge is most apparent given that such signatures must be identified in spite of tremendous inter and intra-subject variability and noise. Crucial for uncovering these signatures has been the development of methods that exploit general statistical properties of natural signals. The signal processing and applied mathematics communities have developed, in recent years, signal representations which take advantage of Gabor-type and wavelet-type functions that localize signal energy in a joint time-frequency and/or space-frequency domain. These techniques can be expressed as multi-resolution transformations, of which perhaps the best known is the wavelet transform. In this paper we review wavelets, and other related multi-resolution transforms, within the context of identifying signatures for disease. These transforms construct a general representation of signals which can be used in detection, diagnosis and treatment monitoring. We present several examples where these transforms are applied to biomedical signal and imaging processing. These include computer-aided diagnosis in mammography, real-time mosaicking of ophthalmic slit-lamp imagery, characterization of heart disease via ultrasound, predicting epileptic seizures and signature analysis of the electroencephalogram, and reconstruction of positron emission tomography data.

Capturing contextual dependencies in medical imagery using hierarchical multi-scale models

Abstract: In this paper we summarize our results for two classes of hierarchical multi-scale models that exploit contextual information for detection of structure in mammographic imagery. The first model, the hierarchical pyramid neural network (HPNN), is a discriminative model which is capable of integrating information either coarse-to-fine or fine-to-coarse for microcalcification and mass detection. The second model, the hierarchical image probability (HIP) model, captures short-range and contextual dependencies through a combination of coarse-to-fine factoring and a set of hidden variables. The HIP model, being a generative model, has broad utility, and we present results for classification, synthesis and compression of mammographic mass images. The two models demonstrate the utility of the hierarchical multi-scale framework for computer assisted detection and diagnosis.