Note to the Reader
A patent for a system to accurately count numbers of cells or particles in a
section or sample of tissue.
1990 Cell Counting Patent
From the Patent and Trademark
Depository Library Program
View GIF images from the patent in a new window:
Front Page |
Columns 1 & 2 |
Columns 3 & 4 |
Columns 5 & 6 |
Columns 7 & 8
Fig. 1 |
Fig. 2 |
Fig. 3 |
Fig. 4a |
|United States Patent
|Williams, et al.
||June 5, 1990
A system is disclosed for counting particles/cells within a counting box
of precisely known volume that is completely inside a transparent section or
sample. The box has a chosen height with defined upper and lower limits and
appropriately selected width and depth dimensions. It resides completely
within the sample and has no surface in common with an exterior surface of
the sample. The system includes a compound light microscope that has a depth
of focus which is small in relation to the thickness dimension of the
counting box. The microscope includes adjustment means for moving the focal
plane through a range which is greater than the height of the counting box.
Display means are provided which show the portion of the sample that is
within the depth of focus and user-operated means is provided to enable the
user to mark the cells so displayed. Indicator means are further provided to
either audibly or visually indicate to the user when the adjustment means
cause the focal plane to pass beyond either the upper or lower height limits
of the counting box. The indicator means further includes means for
accumulating a count of cells within the counting volume as the user
operates the marker means. Means are also provided to compensate for optical
||Williams; Robert W. (Hamden, CT);
Rakic; Pasko (Hamden, CT)
||Yale University (New Haven, CT)
||November 4, 1988
||G01N 033/48; G06M 011/02;
|Field of Search:
U.S. Patent Documents
||Kamachi et al.
Frost, Harold, M.D., Henry Ford Hospital Bulletin, 8, "Measurement of
Osteocytes Per Unit Volume and Volume Components of Osteocytes and
Canaliculae in Man"--pp. 208-211, before 11-4-88.
Abercrombie, M., Anat. Rec. 94, 1946, "Estimation of Nuclear Population
from Microtome Sections"--pp. 239-246.
Petran, M. et al, Journal of the Optical Society of America, vol. 58,
No. 5, May 1968, "Tandem-Scanning Reflected-Light Microscope"--pp.
Padawer, J., Journal of The Royal Microscopic Society, vol. 88, Pt. 3,
Jun. 1968, "The Nomarski interference-contrast microscope, An
experimental basis for image interpretation"--pp. 305-349.
Underwood, E. E., Journal of Microscopy, vol. 89, Pt. 2, Apr. 1969, "Stereology,
or the quantitative evalutaion of microstructures"--pp. 161-180.
Glaser, Edmund M., Journal of Neuroscience Methods, 5(1982), "Snell's
Law: The Bane of Computer Microscopists"--pp. 201-202.
Gunderson et al, Journal of Microscopy, vol. 131, Pt. 1, Jul. 1983,
"Estimation of Section Thickness Unbiased by Cutting-Deformation"--pp.
Howard et al, Journal of Microscopy, vol. 138, Pt. 2, May 1985,
"Unbiased estimation of particle density in the tandem scanning
reflected light microscope"--pp. 203-212.
Curcio et al, Anat. Rec., 1986, "Computer-Assisted Morphometry Using
Video-Mixed Microscopic Images and Computer Graphics"--pp. 329-337.
Gunderson, H. J. G., Journal of Microscopy, vol. 143, Pt. 1, Jul. 1986,
"Stereology of arbitrary particles"--pp. 3-45.
Primary Examiner: Heyman; John S.
Attorney, Agent or Firm: Perman & Green
This invention was made with Government support under Grant numbers ROEY
02593 and PONS 22807 awarded by the Department of Health and Human Services.
The Government has certain rights in the invention.
1. In a system for counting particles within an interior volume of a
transparent sample, said volume having chosen thickness and width
dimensions, said thickness dimension exhibiting upper and lower height
limits, said interior volume containing no surface coincident with an
exterior surface of said sample, the combination comprising:
a compound light microscope with a focal plane whose depth of field is small
in relation to said thickness dimension, said microscope having adjustment
means for moving said sample through a range greater than said thickness
display means for showing particles in said sample which are within said
marker means for a marking a particle to be counted;
indicator means for indicating when said adjustment means causes said focal
plane to overlap an upper or lower height limit; and
means for accumulating a count of marked particles.
2. The system as recite din claim 1 wherein said compound light microscope
is provided with differential interference contrast optics to enhance the
particles which are within said focal plane.
3. The system of claim 1 further comprising:
means for altering the showing of said display means when said adjustment
means is operated to cause said focal plane to overlap an upper or lower
4. The system of claim 3 wherein said altering means modifies said display
means in a first fashion when said overlap of said upper height limit occurs
and modifies said display means in a second fashion when said overlap of
said lower limit occurs.
5. The system of claim 4 wherein said first fashion causes said display
means to exhibit a first color and said second fashion a second color.
6. The system of claim 1 wherein said indicator means comprises alarm means
which provides audible tones when said adjustment means is operated so as to
cause said focal plane to overlap an upper or lower height limit.
7. The system of claim 6 wherein said alarm means provides a first audible
tone when said overlap of said upper limit occurs and a different audible
tone when said overlap of said lower limit occurs.
FIELD OF THE INVENTION
This invention relates to tissue analysis and, more particularly, to an
apparatus for accurately counting the number of cells or particles in a
section or sample of tissue.
BACKGROUND OF THE INVENTION
It is well known to scientists and physicians that the number and size of
cells in a tissue sample is an important factor in determining whether
tissue is healthy or diseased. For instance, the number of cells in parts of
the brain is a fundamental determinant of behavior and cognitive ability.
Even a small deficit or surplus in the number of neurons will have
long-lasting effects on performance (e.g., Parkinson's disease, Huntington's
chorea and perhaps even schizophrenia).
In the prior art there are numerous suggestions of how to count cells or
particles in samples of tissue. One of the more widely used methods was
described by Abercrombie in a paper entitled "Estimation of Nuclear
Populations From Microtome Sections", Anat. Rec., Vol. 94, pp 239-247
(1946). Using Abercrombie's method, sections of tissue are typically cut in
thicknesses of 5 to 20 microns. A counting frame is placed in the eyepiece
of a conventional compound microscope and cells in each section are then
counted at a magnification of 500 or higher. All cells, cell nuclei or
nucleoli that are inside the frame are counted. In addition, those cells
that cross 2 out of 4 edges of the four-sided frame--by convention, the
right and upper edges--are also counted. In contrast, cells that cross any
part of the bottom or left edges of the frame are not counted. The procedure
is repeated from the top to the bottom of the section until all cells at all
depths have been counted. However, during this procedure no account is
actually taken of the three dimensional position of cells, and no attempt is
made to determine the number of cells at the top and bottom surfaces of the
section that were split or dislodged by the knife during cutting.
A more accurate counting method was described by Howard et al in a paper
entitled "Unbiased Estimation Of Particle Density In The Tandem Scanning
Reflected Light Microscope", Journal of Microscopy, Vol. 138, Part II, May,
1985, pp 203-212 (1985). Howard et al. suggest the direct optical
examination of a cube of tissue. They point out that the sample cube is not
physically sectioned but rather that it is examined in an intact state. A
particular cube is chosen and the particles or cells that lie entirely
inside the cube and half of those which transect its surfaces are counted.
In specific, Howard et al. studied unsectioned semiopaque materials (entire
bones) that could not be examined with a normal compound light microscope.
Therefore, Howard et al were constrained to use a tandem scanning reflected
light microscope (TSRLM), an expensive special-purpose microscope that
accepts only directly reflected light. Such an instrument is not designed
for use with conventional sectioned tissue. For a description of the TSRLM,
see "Tandem-Scanning Reflected Light Microscope" by Petran et al, Journal of
the Optical Society of America, Vol. 58, No. 5, May, 1968, pp 661-664.
The method of Howard et al. suffers from several additional drawbacks. It
does not take into account the fact that the number of particles or cells
can actually change during the preparation of a tissue sample. For instance,
if the sample is prepared by microtome cutting, cells/particles can either
be sectioned, pushed deeper into the sample, or pulled out of the sample.
These effects are illustrated in FIG. 1 by cell 10, cell 12 and pit 14,
respectively. If any cube used as suggested by Howard et al. includes such a
section of tissue with altered cells, an erroneous count may occur.
In addition, the method of Howard et al. lacks a convenient way to define
the cubic counting box. The upper and lower surfaces of the counting box are
defined by reference to the microscope's fine-focus scale. The operator must
therefore examine the fine-focus control after each movement to see whether
the limits of the counting box (cube) have been reached. Thus, after each
examination, the operator must glance at the picture, make the particle
count and move the focus control, reexamine the position, and repeat the
process iteratively until the entire depth dimension of the cube has been
Accordingly, it is an object of this invention to adapt a modified Howard et
al. method to use with a conventional compound light microscope and for
examination of typical sectioned materials used in histology, pathology and
It is still another object of this invention to facilitate the operator's
definition of a counting box so that the method is easier to implement.
It is a further object of this invention to provide an improved apparatus
for counting cells in a sample, which avoids inaccuracies introduced by
SUMMARY OF THE INVENTION
A system is disclosed for counting particles or cells within a box of
precisely known volume that is completely inside a transparent section or
sample. The box has a chosen height, with defined upper and lower limits and
appropriately selected width dimensions. The box resides completely within
the sample and has no surface in common with an exterior surface of the
sample. The system includes a compound light microscope that has a depth of
focus which is small in relation to the thickness dimension of the counting
box. The microscope includes adjustment means for moving the focal plane
through a range which is greater than the height of the counting box.
Display means are provided which show the portion of the sample which is
within the depth of focus. User-operated means are provided to enable the
user to mark the cells so displayed. Indicator means are provided to either
audibly or visually indicate to the user when the adjustment means cause the
focal plane to pass beyond either the upper or lower height limits of the
counting box. The indicator means further includes means for accumulating a
count of cells within the counting volume as the user operates the marker
means. Means are also provided to compensate for optical foreshortening.
DESCRIPTION OF THE DRAWINGS
FIG. 1 is a magnified section of tissue showing a counting box in a tissue
FIG. 2 is a diagram that illustrates the major components of the invention.
FIG. 3 is a schematic view of the major optical components of a differential
interference contrast (DIC) optics system.
FIGS. 4a and 4b illustrate a flow diagram showing the procedure used by the
DETAILED DESCRIPTION OF THE INVENTION
Referring now to FIG. 1, a section of tissue 2 is shown which has been
greatly magnified. Tissue section 2 contains a number of cells that are to
be counted. It is well known that tissue sections vary in thickness; (e.g.,
the thickness of a single section can vary by plus or minus 10 microns).
This variation has serious detrimental affects on the accuracy of an
estimate of cell number when conventional methods are used. The reason is
that sample volume will vary in proportion to the variation in section
thickness. This gives rise to very high sampling variations and error. This
error is eliminated in this invention by selecting a counting box 4, no
surface of which is coextensive with an outer surface of the tissue section.
Within counting box 4, a number of cells 16, 18, 20, 22, etc., are found. In
accordance with the method of Howard et al., only those cells that transect
three surfaces of counting box 4 are counted, whereas those that transect
the other three surfaces are ignored. For instance, cells transecting
surfaces that are bounded by a dotted line in FIG. 1 would be ignored, while
cells that transect surfaces bounded solely by solid lines would be counted.
Above and below counting box 4 are buffer zones 30 and 32, which should,
ideally, be at least as high as the tallest cell/particle to be counted.
Buffer zones 30 and 32 isolate counting box 4 from surface artifacts of all
Referring to FIG. 2, the major components that comprise the invention will
be described. Microscope 50 is a conventional compound light microscope. It
is preferably provided with a high-power objective having a narrow depth of
field. As will be described below, microscope 50 is further preferably
provided with differential interference contrast (DIC) optics to enhance the
user's ability to identify particular cell structures.
A video camera 52 is affixed, in the normal manner, to the upper portion of
microscope 50 and feeds its signal to video processor 54. Video processor 54
is provided with a synchronizing input 56 from a microcomputer 58. Video
processor 54 has controls that enable the operator to adjust and optimize
brightness, contrast, and resolution of the image. The output signal from
video processor 54 is fed via line 60 to a video mixer 62. Another input to
video mixer 62 is provided via line 64 from microcomputer 58. Inputs on line
64 enable the line segments of a counting frame to be superimposed on the
scene being transmitted from video processor 54. The counting frame is
defined by lines 33a, 33b, 33c and 33d in FIG. 1.
The output of video mixer 62 is applied to and viewed on a video monitor 66,
which preferably has the ability to provide color images to the viewer.
Monitor 66 is of the "RGB" type, which enables its color presentation to be
automatically controlled by signals from microcomputer 58 (i.e., its red,
blue and green controls are available for individualized control). An input
device 68 (e.g., mouse, keyboard) is connected to microcomputer 58 and
provides the operator with the ability to construct the counting frame as
well as to mark and count the cells shown on monitor 66.
As will be hereinafter understood, it is important that the height of the
counting box and sample be measured with accuracy. However, it is known that
the apparent height of objects examined with a microscope is a function of
differences in the refractive index of the tissue, the objective, and the
materials in which the tissue and slide are immersed. If a dry objective is
used to focus first on the top and then on the bottom of an object, the
distance the stage actually travels will only be 66% of the true distance
between top and bottom. The difference is accounted for by the ratio of the
refractive indices of air and crown glass (1.000:1.524). An object 45
microns high will be imaged over a vertical distance of only 30 microns.
Obviously, therefore, the refraction of light is a problem when the operator
tries to define the height of a counting box. Z axis comparator 73 can
compensate for optical foreshortening. The operator can readily enter a
correction factor that can be applied automatically.
Another solution is to use an oil immersion objective. With homogeneous oil
immersion, the tissue, the materials in which the tissue is placed, and the
glass of the objective have refractive indices close to each other.
Immersion oil has a refractive index of 1.515. The mounting material (e.g.,
Fisher's Permount) has a refractive index of 1.515 (wet) or 1.529 (dry).
Cellulose nitrate embedding material has a refractive index of 1.514. Cover
glass has a refractive index between 1.513 and 1.534. These indices are so
close to each other that the optical and the true height of an object
correspond almost exactly.
Microscope 50 is preferably modified in several ways. First, it is provided
with a digital micrometer 70, which provides outputs on line 72 to Z axis
comparator 73 indicative of the movements of the stage 76. The position of
the focal plane of microscope 50 is indicated by such outputs. Z axis
comparator 73 is provided with manual inputs (not shown) which enable set
points to be entered indicating the upper and lower height thresholds of the
counting box. Those set points may also be entered from microcomputer 58 via
As is well known, the relative position of stage 76 may be varied by an
operator's movement of the fine focus control 78. It should be understood
that other position sensing devices may be used in lieu of digital
micrometer 70. For instance, a shaft encoder can be attached to the fine
focus control 78 to provide an input to Z axis comparator 73. However, a
digital micrometer that directly senses movements of the stage is preferable
because the measurements are direct. One micrometer which has been employed
is the Metro-25 Digital Length Gauge, which has a range of movement of 25
millimeters, a resolution of 0.1 microns, and an accuracy of better than 0.5
microns over its full range. That micrometer was obtained from Heidenham
Corporation, Elk Grove Village, Ill.
A second modification to the microscope involves the inclusion therein of
differential interference contrast (DIC) optics. Such optics are also known
as Nomarski interference-contrast optics, and this system is described by
Padawer in his paper entitled "The Nomarski Interference-Contrast
Microscope, An Experimental Basis for Image Interpretation," Journal of The
Royal Microscope Society, Vol. 88, Part 3, June 1968 pp. 305-349. FIG. 3
shows a schematic of the optics contained in a DIC optical system.
The reason for inclusion of DIC optics is that it is often difficult to
obtain high-contrast images with conventional microscope optics. The reason
is that the image is partly obscured by the blur of structures just outside
the microscope's plane of focus. DIC optics remove low-spatial-frequency
blur while enhancing high-spatial-frequency detail. Furthermore, they tend
to enhance the definition of the equatorial portion o a cell's outline
(widest diameter portion of the cell). DIC images are therefore easier to
focus and the operator finds it easier to decide whether the edge of a cell
or particle extends beyond the boundaries of a counting box.
The DIC optical system shown in FIG. 3 may be characterized as a
polarized-light interferometer. The light source is incoherent. The light
traverses polarizer P1 before entering a compensator prism Qc. Polarizer P1
selects a set of light waves vibrating at +45 degrees. The light polarized
at +45 degrees enters a modified Wollaston prism or birefringent
beamsplitter Qo which gives rise to two orthogonally polarized wave fronts.
These emerging waves are offset by a variable amount as they pass through
the tissue. When these two wave fronts are brought back to the same plane
above the objective, they interfere with each other. The interference
between them is rendered observable when they pass through analyzer P2. This
interference phenomenon substantially enhances the outlines of cells and/or
particles for the microscope operator.
The operation of the system shown in FIG. 2 will be understood-by reference
to FIGS. 4a and 4b. As shown in box 100 of FIG. 4a, the operator initially
measures the thickness of the sample section. This is done by initially
focusing on the uppermost surface of the sample and then racking down the
objective until the lowermost surface is in focus. If Z axis comparator 73
(FIG. 2) is zeroed at the uppermost surface, it will indicate the thickness
of the sample when the lowermost portion is in focus.
Knowing the thickness of the sample section, the operator selects the
position and dimensions of the counting box (box 102). If the sample section
is approximately 40 microns thick, the lower surface of the counting volume
will be set approximately 3 microns above the bottom of the section and the
upper surface approximately 28 microns above the bottom of the section. This
gives a counting box 25 microns high. In a similar fashion, the operator
chooses the length and the width of the counting frame (equivalent to the
length and width of the counting box).
Subsequently, the operator enters into microcomputer 58 the top height limit
of the counting box (box 104), the bottom height limit of the counting box
(box 106), and the coordinates of the line segments that define the length
and width of the counting frame (box 108). The operator enters the top and
bottom height limits by inputting them through the keyboard of microcomputer
58. The operator enters the counting box line segments by using mouse 68 or
another appropriate input device to draw orthogonal lines on display 66 and
those lines define the desired length and width dimensions of the counting
frame for microcomputer 58.
When the upper and lower height limits are entered into microcomputer 58,
they are fed via conductor 80 (FIG. 2) to Z axis comparator 73. When it
receives a signal via conductor 72 that stage 76 has reached a top or bottom
height limit, Z axis comparator provides two types of outputs, each
selectable by an operator. The first selectable output is an audible. alarm.
that a height limit has been reached. It is preferred that the audible alarm
have two distinct tones, one for the top height limit and one for the bottom
limit. This enables the operator to easily distinguish, by ear, that the
stage has reached an operating limit, without the need to constantly monitor
the height readout.
The second type of selectable output from Z axis comparator 73 is a change
in the color display on monitor 66 when a height limit has been reached.
This output is sent via conductor 80 to microcomputer 58, which then
inhibits a color input to monitor 66. Thus, as the operator moves stage 76,
the color displayed on monitor 66 changes, thereby indicating when a top or
bottom height limit of the counting box has been reached or exceeded.
Preferably, one component color (e.g. red) is inhibited when passing through
the bottom height limit, and a different color (e.g., blue) is inhibited
when passing through the top height limit. The audible alarms and color
indications may be used either separately or together.
Returning to FIG. 4a, the operator moves stage 76 to the bottom limit of the
counting box (box 110) and then begins moving the stage upward. The operator
marks and counts each cell as it comes into focus on monitor 66. The
operator then uses input device 68 (box 114) to trace the outline of each
in-focus cell that is partially or completely within the counting frame. In
addition to keeping track of the cell tracings, microcomputer 58 keeps a
count of the number of tracings (box 116). It should be remembered that
cells that intersect the bottom limit are not counted. For that reason, any
cell that intersects the bottom height limit is not traced or counted. These
uncounted cells are recognized and rejected by commencing the stage movement
below but near the bottom limit. Cells that are visible on both sides of the
bottom limit are rejected.
If the upper limit of the counting box has not been reached, the procedure
continues: the operator again moves the stage upward until additional cells
are in focus, and then marks and then marks and counts them (decision box
118). Once the upper height limit of the counting volume is reached (as
indicated to the operator by an audible alarm and/or a color change on
monitor 66), the procedure branches. The operator instructs microcomputer 58
to count all cell tracings within the counting box that do not intersect a
counting frame line segment (box 120). Thus, all cells whose margins lie
entirely within the counting frame are counted in a cumulative manner and
the value is stored. Next, the operator instructs microcomputer 58 to
subtract from the count all cell tracings that intersect the line segments
that define the right and top edges of the counting frame. (Cells that
intersect the left and bottom edges of the counting frame are included in
the count.) Microcomputer 58 accomplishes this by determining whether any
solutions of the line equation for a particular counting-frame line segment
lie within the limits of a cell tracing. If such an overlap is found, then
it is known that the cell tracing intersects a frame line segment and is to
be handled accordingly. Finally, all cells or particles whose margins lie
completely inside the counting box and all cells or particles which
intersect three of the six surfaces of the counting box are added to provide
a complete cell count (box 124).
This process is repeated at different locations within the tissue sample and
the results are combined. Subsequently, the operator can obtain an estimate
of the total number of cells or particles in the tissue sample by
multiplying the mean density by the total volume of the region in which the
counting boxes have been located.
It should be understood that the foregoing description is only illustrative
of the invention. Various alternatives and modifications can be devised by
those skilled in the art without departing from the invention. Accordingly,
the present invention is intended to embrace all such alternatives,
modifications, and variances which fall within the scope of the appended
* * * * *