Mark Schulze

Mathematical morphology is a set algebra that defines some important new techniques in image processing. Morphological filters are closely related to order statistic and other nonlinear filters, but they are uniquely sensitive to shape. A morphological filter will preserve shapes similar to its structuring element shape while modifying dissimilar shapes. Most morphological filters are effective at removing both linear and nonlinear noise processes. However, the standard morphological operators introduce a statistical and deterministic bias to images. Fortunately, these operators exist in complementary pairs that are equally and oppositely biased. One way to alleviate the bias is to average the two complementary operators. The filters formed by such averages are the midrange filter (basic operators), the pseudomedian filter (singly compound operators) and the LOCO filter (doubly compound operators). In thermographic imaging, one often wishes to find exact temperatures or accurate isothermal contours. Therefore, techniques used to remove sensor noise and scanning artifact should not introduce bias. The LOCO filter that we have devised provides the shape control and noise suppression of morphological techniques without biasing the image. We will demonstrate the effects of different structuring element shapes on thermographic images of tissue heated by laser irradiation and electrosurgery.

ABSTRACT Fluorescence in situ hybridization (FISH) of metaphase chromosome spreads is valuable for monitoring the radiation dose to circulating lymphocytes. At low dose levels, the number of cells that must be examined to estimate aberration frequencies is quite large. An automated microscope that can perform this analysis autonomously on suitably prepared specimens promises to make practical the large-scale studies that will be required for biodosimetry in the future. This paper describes such an instrument that is currently under development. We use metaphase specimens in which the five largest chromosomes have been hybridized with different-colored whole-chromosome painting probes. An automated multiband fluorescence microscope locates the spreads and counts the number of chromosome components of each color. Digital image analysis is used to locate and isolate the cells, count chromosome components, and estimate the proportions of abnormal cells. Cells exhibiting more than two chromosomal fragments in any color correspond to a clastogenic event. These automatically derived counts are corrected for statistical bias and used to estimate the overall rate of chromosome breakage. Overlap of fluorophore emission spectra prohibits isolation of the different chromosomes into separate color channels. Image processing effectively isolates each fluorophore to a single monochrome image, simplifying the task of counting chromosome fragments and reducing the error in the algorithm. Using proportion estimation, we remove the bias introduced by counting errors, leaving accuracy restricted by sample size considerations alone.

ABSTRACT The pseudomedian filter was designed to be a computationally efficient alternative to the median filter. The output of the pseudomedian filter is the average of two values obtained by maximum and minimum operations performed on selected subwindows within the filter window. Although the response of the pseudomedian filter resembles that of the median filter in many ways, there are important differences. The pseudomedian filter is more susceptible to impulse-like noise in images, but the square-shaped 2D pseudomedian filter does not round off sharp corners as much as the square-shaped 2D median filter does. The psuedomedian filter also removes high- frequency periodic elements from images. The pseudomedian filter exhibits a more 'center-weighted' response than the median filter. Thus, fine details that are completely removed by the median filter usually remain visible in pseudomedian- filtered images. Images filtered by a square-shaped 2D pseudomedian filter often have a 'blocky' appearance caused by the square shape of the filter subwindows. These properties of the pseudomedian filter indicate that it may be more appropriate than the median filter for processing images with sharp corners or fine details.

ABSTRACT In multiscale edge detection we exploit both the high locational accuracy of edges detected at fine scale and the robustness to noise at coarse scales, obtaining a final edge map possessing both of these desirable characteristics. In the proposed centroid attraction technique a coarse scale edge map is initially generated, and the locations of its edge elements are corrected by 'attraction' to the edge elements of progressively finer scaled edge maps. Centroid attraction is demonstrated here for the challenging application of synthetic aperture radar (SAR) imagery with its high amplitude multiplicative noise, using both real and synthesised examples.

A developmental instrument for assessment of radiation-induced damage in human lymphocytes includes an automated fluorescence microscope equipped with a one or more chargecoupled- device (CCD) video camera(s) and circuitry to digitize the video output. The microscope is also equipped with a three-axis translation stage that includes a rotation stage, and a rotary tray that holds as many as thirty specimen slides. The figure depicts one version of the instrument. Once the slides have been prepared and loaded into the tray, the instrument can operate unattended. A computer controls the operation of the stage, tray, and microscope, and processes the digital fluorescence-image data to recognize and count chromosomes that have been broken, presumably by radiation. The design and method of operation of the instrument exploit fluorescence in situ hybridization (FISH) of metaphase chromosome spreads, which is a technique that has been found to be valuable for monitoring the radiation dose to circulating lymphocytes. In the specific FISH protocol used to prepare specimens for this instrument, metaphase lymphocyte cultures are chosen for high mitotic index and highly condensed chromosomes, then several of the largest chromosomes are labeled with three of four differently colored whole-chromosome-staining dyes. The three dyes, which are used both individually and in various combinations, are fluorescein isothiocyanate (FITC), Texas Red (or equivalent), and Cy5 (or equivalent); The fourth dye 4',6-diamidino- 2-phenylindole (DAPI) is used as a counterstain. Under control by the computer, the microscope is automatically focused on the cells and each slide is scanned while the computer analyzes the DAPI-fluorescence images to find the metaphases. Each metaphase field is recentered in the field of view and refocused. Then a four-color image (more precisely, a set of images of the same view in the fluorescent colors of the four dyes) is acquired. By use of pattern-recognition software developed specifically for this instrument, the images in the various colors are processed to recognize the metaphases and count the chromosome fragments of each color within the metaphases. The intermediate results are then further processed to estimate the proportion of cells that have suffered genetic damage. The prototype instrument scans at an average areal rate of 4.7 mm2/h in unattended operation, finding about 14 metaphases per hour. The false-alarm rate is typically less than 3 percent, and the metaphase-miss rate has been estimated to be less than 5 percent. The counts of chromosomes and fragments thereof are 50 to 70 percent accurate.

In multiscale edge detection we exploit both the high locational accuracy of edges detected at fine scale and the robustness to noise at coarse scales, obtaining a final edge map possessing both of these desirable characteristics. In the proposed centroid attraction technique a coarse scale edge map is initially generated, and the locations of its edge elements are corrected by 'attraction' to the edge elements of progressively finer scaled edge maps. Centroid attraction is demonstrated here for the challenging application of synthetic aperture radar (SAR) imagery with its high amplitude multiplicative noise, using both real and synthesised examples.

Mathematical morphology is a set algebra that defines some important new techniques in image processing. Morphological filters are closely related to order statistic and other nonlinear filters, but they are uniquely sensitive to shape. A morphological filter will preserve shapes similar to its structuring element shape while modifying dissimilar shapes. Most morphological filters are effective at removing both linear and nonlinear noise processes. However, the standard morphological operators introduce a statistical and deterministic bias to images. Fortunately, these operators exist in complementary pairs that are equally and oppositely biased. One way to alleviate the bias is to average the two complementary operators. The filters formed by such averages are the midrange filter (basic operators), the pseudomedian filter (singly compound operators) and the LOCO filter (doubly compound operators). In thermographic imaging, one often wishes to find exact temperatures or accurate iso...

23 A developmental instrument for assessment of radiation-induced damage in human lymphocytes includes an automated fluorescence microscope equipped with a one or more chargecoupled-device (CCD) video camera(s) and circuitry to digitize the video output. The microscope is also equipped with a three-axis translation stage that includes a rotation stage, and a rotary tray that holds as many as thirty specimen slides. The figure depicts one version of the instrument. Once the slides have been prepared and loaded into the tray, the instrument can operate unattended. A computer controls the operation of the stage, tray, and microscope, and processes the digital fluorescence-image data to recognize and count chromosomes that have been broken, presumably by radiation. The design and method of operation of the instrument exploit fluorescence in situ hybridization (FISH) of metaphase chromosome spreads, which is a technique that has been found to be valuable for monitoring the radiation dose to circulating lymphocytes. In the specific FISH protocol used to prepare specimens for this instrument, metaphase lymphocyte cultures are chosen for high mitotic index and highly condensed chromosomes, then several of the largest chromosomes are labeled with three of four differently colored whole-chromosome-staining dyes. The three dyes, which are used both individually and in various combinations, are fluorescein isothiocyanate (FITC), Texas Red (or equivalent), and Cy5 (or equivalent); The fourth dye — 4',6-diamidino2-phenylindole (DAPI) — is used as a counterstain. Under control by the computer, the microscope is automatically focused on the cells and each slide is scanned while the computer analyzes the DAPI-fluorescence images to find the metaphases. Each metaphase field is recentered in the field of view and refocused. Then a four-color image (more precisely, a set of images of the same view in the fluorescent colors of the four dyes) is acquired. By use of pattern-recognition software developed specifically for this instrument, the images in the various colors are processed to recognize the metaphases and count the chromosome fragments of each color within the metaphases. The intermediate results are then further processed to estimate the proportion of cells that have suffered genetic damage. The prototype instrument scans at an average areal rate of 4.7 mm/h in unattended operation, finding about 14 metaphases per hour. The false-alarm rate is typically less than 3 percent, and the metaphase-miss rate has been estimated to be less than 5 percent. The counts of chromosomes and fragments thereof are 50 to 70 percent accurate. This work was done by Kenneth R. Castleman and Mark Schulze of Perceptive Scientific Instruments, Inc., for Johnson Space Center. For further information, contact the Johnson Technology Transfer Office at (281) 4833809. MSC-23072 Radiation Dosimetry via Automated Fluorescence Microscopy With further development, this instrument could enable biodosimetry on a large scale. Lyndon B. Johnson Space Center, Houston, Texas Bio-Medical

Non-vibrating-way contact potential difference probe and the vibration method using a contact potential difference probe, a method and system for determining the contact potential difference of the wafer surface is disclosed. The method and system non-vibrating-way contact potential difference between the wafer surface scanning by the sensor, the integration and scaling, and the integrated and scaled data of the last data vibration-individual in order to match to the obtained measurements using the method contact potential difference sensor It shall involve the application of an offset to the data track.

The discriminatory power and imaging efficiency of different multicolor FISH (M-FISH) analysis systems are key factors in obtaining accurate and reproducible classification results. In a recent paper, Garini et al. put forth an analytical technique to quantify the discriminatory power ("S/N ratio") and imaging efficiency ('excitation efficiency') of multicolor fluorescent karyotyping systems. A parametric model of multicolor fluorescence microscopy, based on the Beer-Lambert law, is analyzed and reduced to a simple expression for S/N ratio. Parameters for individual system configurations are then plugged into the model for comparison purposes. We found that several invalid assumptions, which are used to reduce the complex mathematics of the Beer-Lambert law to a simple S/N ratio, result in some completely misleading conclusions about classification accuracy. The authors omit the most significant noise source, and consider only one highly abstract and unrepresentative situation. Unwisely chosen parameters used in the examples lead to predictions that are not consistent with actual results. The earlier paper presents an inaccurate view of the M-FISH situation. In this short communication, we point out several inaccurate assumptions in the mathematical development of Garini et al. and the poor choices of parameters in their examples. We show results obtained with different imaging systems that indicate that reliable and comparable results are obtained if the metaphase samples are well-hybridized. We also conclude that so-called biochemical noise, not photon noise, is the primary factor that limits pixel classification accuracy, given reasonable exposure times.

The pseudomedian filter was designed to be a computationally efficient alternative to the median filter. The output of the pseudomedian filter is the average of two values obtained by maximum and minimum operations performed on selected subwindows within the filter window. Although the response of the pseudomedian filter resembles that of the median filter in many ways, there are important differences. The pseudomedian filter is more susceptible to impulse-like noise in images, but the square-shaped 2D pseudomedian filter does not round off sharp corners as much as the square-shaped 2D median filter does. The psuedomedian filter also removes high- frequency periodic elements from images. The pseudomedian filter exhibits a more 'center-weighted' response than the median filter. Thus, fine details that are completely removed by the median filter usually remain visible in pseudomedian- filtered images. Images filtered by a square-shaped 2D pseudomedian filter often have a 'blocky' appearance caused by the square shape of the filter subwindows. These properties of the pseudomedian filter indicate that it may be more appropriate than the median filter for processing images with sharp corners or fine details.