Dynamic images are temporal sequences of images, where the intensities of certain regions of interest (ROI's) change with time, whereas anatomical structures remain stationary. Here, new applications of dynamic image analysis, called similarity mapping, are reviewed. Similarity mapping identifies regions in a dynamic image sequence according to their temporal similarity or dissimilarity with respect to a reference ROI. Pixels in the resulting similarity map whose temporal sequence is similar to the reference ROI have high correlation values and are bright, while those with low correlation values are dark. Therefore, similarity mapping segments structures in a dynamic image sequence based on their temporal responses rather than spatial properties. The authors describe the abilities of similarity mapping to identify different image structures present in several dynamic MRI datasets with potential clinical value. They demonstrate that similarity mapping technique has been successful in identifying the following structures: 1) renal cortex and medulla, 2) activated areas of the brain during photic stimulation, 3) ischemia in the left coronary artery territory, 4) lung tumor, 5) tentorial meningioma, and 6) a region of focal ischemia in brain.

Three-dimensional mathematical morphology and its implementation in three-dimensional cellular automation are discussed. Automata having up to 262,144 PEs (processing elements) have been emulated using the Triakis software. Each PE in addition to its present state (either on or off) is permitted by Triakis software to store up to 20 previous states and is assigned a program word having 2/sup N/ binary locations (bits), where (N-1) is the number of neighbors to which the PE is connected. Triakis uses the FCC (face-centered cubic) tessellation, where the kernel is the tetradekahedron and N=13. The problem of embedding the tetradekahedron in the 64*64*64 array is addressed, and properties of the FCC tessellation are examined. The use of the approach for analyzing true three-dimensional binary data from CT (computer tomography) and MR (magnetic resonance) and three-dimensional binary arrays generated from column encoding gray-level imagery is reported.>

The cellular logic transform has been used extensively in the analysis of bilevel images generated by thresholding gray level images. Now, by a process of gray level resynthesis, it is shown to be useful in gray level image processing.

This correspondence reviews the logical transform and pre-sents new theoretical and empirical developments. Novel extensions of the logical transform into spaces of three and four dimensions are de-scribed using the hexahedral and octahedral tessellations. These new tessellations are important in that they yield sufficiently compact neighborhoods to permit multidimensional cellular logic transforms to be carried out in real time by table lookup. Applications of logical transforms in these tessellations to data and image analysis are provided.

A new logical transform processor has been designed and its performance presented in this correspondence. The design incorporates the speed advantage of a locally parallel processor and yet has the capability of processing images of widely ranging dimension. Sixteen pixels are processed simultaneously using table lookup. Image size and neighborhood size are under program control of the host computer, a Perkin-Elmer 3220. This logical transform processor is capable of executing picture point operations in 50 ns.

Cellular logic machines used for feature extraction in pattern recognition have increased in speed to the point of making it possible to execute programs equivalent to 1 billion general-purpose computer instructions in 1 TV frame time. Unfortunately, most cellular logic operators (CLO's) are designed ad hoc. It is important, therefore, to begin to systematize the generation of algorithms using CLO sequences for pattern analysis. These notes systematically analyze some aspects of CLO's which are used in shape discrimination and idealization and in object counting and sizing. New extensions of subfield numbering schemes in the hexagonal tessellation are introduced.

It is estimated that approximately fifty image processing languages are in common use worldwide. Those languages which are specific to biomedical image analysis are the subject of this paper. In general, image processing languages range from those which are self-contained, i.e., include their own monitor, parser, interpretor, memory management, and input/output routines, to those which are simply collections of FORTRAN subroutines called by a command language using the command substitution system present in most modern computer operating systems. Biomedical image analysis systems are divided into (1) those which are available commercially for clin ical image processing and (2) those which are designed for research purposes, concentrating especially, on the analysis of images of cells and cellular architecture in tissues and, fi nally, in radiographic images produced by x-ray equipment and equipment for use in generating computerized tomograms.

A computer program has been written which is capable of evaluating all possible Golay marking transformations related to a particular pattern recognition task using an exhaustive search technique. Such evaluations are reported for separating two image data sets: one taken from biomedical microscopy; the other, aerial reconnaissance.

Golay hexagonal pattern transforms are position independent local operators for use in transforming or altering binary images. The hexagonal tessellation is preferred because it removes the connectivity ambiguity present in the square or checkerboard tessellation. Golay transforms also may be applied to multilevel or "gray" images by encoding such images as a registered stack of binary image planes. The general Golay transform creates a new binary image (the output image) from as many as three stacked input images. Simpler Golay transforms merely alter the binary pattern contained in a single image plane, i.e., the same plane acts as both input and output. Because it is slow and cumbersome to perform Golay transforms using a general-purpose computer, fast special-purpose computers have been built for this purpose which may be programmed in a new image processing language called Glol (Golay logic language). It has been found, using both synthetic images as well as images taken from the real world, that Golay transforms are useful in feature enhancement and extraction. Several illustrative examples are provided.