SOFTWARE FOR IMAGE SEGMENTATION AND ANALYSIS
IN PATHOLOGY (ISAP).
Copyright © 1994, G. William Moore, Geoffrey W. Moore,
Jules J. Berman, Lawrence A. Brown.
http://www.netautopsy.org/ascpisap.htm
U. S. Government Work, presented at the 1994 meeting
of the American Society of Clinical Pathology, Washington, DC.
Moore GW, Berman JJ, Moore GW, Brown LA.
Software for Image Segmentation and Analysis in Pathology (ISAP)
Am J Clin Pathol 102:538-539, 1994.
1. ABSTRACT.
1. Quantitative and qualitative analysis of anatomic pathology images
is currently dominated by commercial software with proprietary
algorithms.
2. Since algorithms are secret, the pathologist cannot necessarily
validate measurements produced by this commercial software.
3. ISAP: Image Segmentation and Analysis in Pathology, is a
public-domain image software application, developed at the Baltimore
VA Medical Center.
4. Pathologists can inspect, improve, and compare available source
code and measurements from ISAP against commercial systems.
2. INTRODUCTION.
1. ISAP: Image Segmentation and Analysis in Pathology, is a
public-domain image software application, which segments an
image into smaller objects and measures these objects.
2. Image segmentation involves dividing the object into smaller
objects, such as cells, nuclei, etc.
3. Objects are measured for features of size, shape, and texture.
4. Results may be viewed on-screen for individual cells, or are
stored consecutively in import format for popular spreadsheets
(Quattro, Excel, Lotus), for further analysis.
3. ISAP SOFTWARE LICENSE.
1. ISAP is copyright 1994, by G. William Moore, Geoffrey W. Moore,
Jules J. Berman, Lawrence A. Brown (hereinafter called the authors),
is prototype software for image segmentation and analysis
in pathology.
2. This software license is the only agreement between users
of ISAP and the authors.
3. There are no implied warranties that are not written in this
software license.
4. There are no liabilities on the authors resulting from the use
or misuse of ISAP.
5. ISAP is not a diagnostic device, and there are no claims implied
or stated concerning its value.
6. The authors are not responsible for any software errors in ISAP
and are not obligated to correct defects in ISAP.
4. MATERIALS AND METHODS.
1. Source code written entirely in Microsoft (R) Visual Basic
for windows, version 3, and provided with the demonstration disk.
2. Accepts uncompressed black-and-white images in targa (.TGA)
or Windows (.BMP) file formats.
3. Operates on IBM PC or compatible computers, with 5 Mb ram memory,
2 Mb available hard disk, MS-DOS version at least 5.0,
Microsoft (R) Windows version 3.1, 95, 98, or NT,
Microsoft (R) file VBRUN300.DLL, and a 256 color display monitor.
4. Results are stored consecutively in comma-delimited import format
for popular spreadsheets (Quattro, Excel, Lotus),
for further analysis.
5. EDGE DETECTION.
1. User makes a tracing to delineate a workspace, which includes
the object (nucleus) and a surrounding rim of cytoplasm and
background.
2. User may preselect a grayvalue-threshold between 0 and 255.
3. User may preselect a percentage-threshold, where x% is inside
the object and (100-x)% is outside the object.
4. For automated threshold, ISAP makes a histogram of all pixels
in the workspace. at each division point between two consecutive
bars of the histogram, ISAP calculates the pooled standard deviation
for the Student t test. minimum pooled standard deviation determines
the automated threshold.
6. VALUES OBTAINED BY ISAP.
1. Size of object: area (area); sum of gray values in the object
(graysum); sum of optical densities in the object (densesum);
average optical density (denseavg); average diameter of the object
(avgdiam); greatest diameter (maxdiam); least diameter (mindiam).
2. Shape of object: elliptical eccentricity (eccent); object contour
(contour); fractal dimension (fractal); perimeter (perimeter).
3. Texture of object: standard deviation of optical density
(densestdv).
7. ISAP ALGORITHMS.
1. Area is the number of pixels inside the object. graysum is the
sum of gray_values inside the object. from each gray_value,
optical_density for that point is calculated by the formula:
optical_density=log10(255/(255-gray_value)), with optical_density
equal zero if gray_value equals 255. Densesum is the sum of
optical_density values inside the object. Denseavg equals
densesum/area. Densestdv is the standard deviation of optical_density.
2. Maxdiam for the object is the Euclidean distance between the two
most distant edge-points. if the object is assumed to be an ellipse,
then: Area = (pi/4) * Maxdiam * Mindiam, and
eccent = Maxdiam/Mindiam. if the object is assumed to be a circle,
then: Area = (pi/4) * Avgdiam * Avgdiam.
3. Perimeter is the sum of edge-points of the object. The object
contour equals perimeter/sqrt(area). Fractal dimension, s, of
an object is a quantitative measure of complexity along the particle
edge. s = 1 + lim(d->0) log2((perimeter at 2*d)/(perimeter at d)),
for gridsize=d, as gridsize approaches 0.
8. FIGURE LEGENDS.
FIGURE 1. Demonstration image collected from
a fine-needle cytologic aspiration of the neck in a 71 year old male
with lymphoma, Papanicolaou staining, 756 by 456 pixel uncompressed 8-bit
black-and-white image.
FIGURE 2. Demonstration image in which several cell nuclei have been
approximately outlined by hand (dark line), and then automatically
thresholded and segmented by the ISAP software.
FIGURE 3. Demonstration image in which a cell nucleus has been
automatically thresholded and segmented by the ISAP software.
Values for nuclear size, shape, and texture are displayed.
FIGURE 4. Demonstration image in which a cell nucleus has been
automatically thresholded and segmented by the ISAP software.
Histogram of grayvalues is displayed. Center line is the arithmetic mean.
FIGURE 5. Demonstration image in which measurements on several cell
nuclei have been collected in a comma-delimited database, for further
analysis in popular spreadsheets (Quattro, Excel, Lotus).
9. CONCLUSION.
1. Analysis of anatomic pathology images currently uses commercial
software with proprietary algorithms, which cannot necessarily be
validated.
2. ISAP: image segmentation and analysis in pathology, is a public-
domain image software application, which may be distributed freely
but cannot be sold.
3. Pathologists can inspect, improve, and compare available source
code and measurements from ISAP against commercial systems.
4. ISAP can serve as a minimum standard for performance
by commercial systems.
10. REFERENCES.
1.
Moore GW, Berman JJ, Moore GW, Brown LA.
Software for Image Segmentation and Analysis in Pathology (ISAP)
Am J Clin Pathol 102:538-539, 1994.
http://www.netautopsy.org/ascpisap.htm
2.
Moore GW, Berman JJ, Moore GW, Brown LA.
Welcome to ISAP.
http://www.medparse.com/isapwlcm.htm
3.
Moore GW, Berman JJ, Moore GW, Brown LA.
ISAP Edge Detection Menu.
http://www.medparse.com/isapedge.htm
4.
Moore GW, Berman JJ, Moore GW, Brown LA.
ISAP Histogram Menu.
http://www.medparse.com/isaphist.htm
5.
Moore GW, Berman JJ, Moore GW, Brown LA.
ISAP Values Menu.
http://www.medparse.com/isapvalu.htm
6.
Moore GW, Berman JJ, Moore GW, Brown LA.
ISAP Troubleshooting Menu.
http://www.medparse.com/isaptrub.htm
7.
Moore GW, Berman JJ, Moore GW, Brown LA.
ISAP Microsoft(R) Visual Basic Source Code.
http://www.medparse.com/isapsour.htm
8.
Moore GW, Berman JJ, Moore GW, Brown LA.
ISAP Formulas.
http://www.medparse.com/isapform.htm
9.
Moore GW, Berman JJ, Moore GW, Brown LA.
ISAP demonstration program and Perl source code.
http://www.medparse.com/isapver2.htm
Last Updated: January 26, 2003, by G. William Moore, MD, PhD.