Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02389271 2002-04-24
WO 01/33493 PCT/US99/25360
1
SYSTEM FOR MEASURING TISSUE SIZE AND MARBLING IN AN ANIMAL
TECHNICAL FIELD
This invention relates to animal measuring systems and more particularly to
measuring
skeletal, intramuscular fat, back fat and muscle tissue contained within a
live animal or carcass.
Even more particularly, the invention relates to measuring skeletal,
intramuscular fat, back fat and
muscle tissue through Magnetic Resonance Imaging systems.
BACKGROUND OF THE INVENTION
People have always made visual appraisals of domestic animals and humans to
compare like
kinds and to try to predict future performance and production. In domestic
animals it is also
beneficial to select a young offspring that will produce a superior animal.
Animal breeders
continually try to select for faster or stronger horses, increased volumes of
meat for cattle, swine,
poultry and sheep as well as a larger volume of milk for dairy animals. The
very economic base of
pricing for animals is directly related to a predicted future performance or
production of the
animals.
It is well known that the highest price is paid for the butchered beef
carcasses that not only
possess the greatest quantity of meat but also the highest percentage of
intramuscular fat, which is
often referred to as "marbling". The United States Department of Agriculture
(U.S.D.A.) uses a
grading system to compare like kinds of meat. The grading system within the
beef industry denotes
the highest quality meat with a rating of U.S.D.A. PRIME. Respectively the
next two ratings are
CHOICE and SELECT. PRIME meat brings the highest price per pound. The other
end of the
spectrum has the lowest ratings of CUTTER and CANNER and bring the lowest
price per pound.
The finest steaks (and highest priced) are often purchased by restaurants and
promoted as U.S.D.A.
PRIME Beef. Many of the steaks purchased in a meat market are graded U.S.D.A.
CHOICE or
SELECT.
When a beef animal is butchered and the actual quantity and quality of meat
inside can be
seen, then the grading and pricing can be very accurate. However, there is a
tremendous need to
determine the potential quantity and quality of the meat when the animal is
younger, many months
prior to butchering. A beef animal is often sold on several occasions
throughout its life before it
is butchered. It may be sold as a weanling (just weaned from its mother cow)
and then may be sold
again months later to a feed lot. In the feed lot the beef animal is given a
concentrated ration of
CA 02389271 2002-04-24
WO 01/33493 PCT/US99/25360
2
food to maximize the growth process as well as maximize the marbling within
the meat. Finally,
the beef animal is marketed to a butchering facility to provide steaks, roast,
hamburgers and many
other beef products.
The beef animal is usually weighed at each point of sale and often
subjectively appraised
by a person knowledgeable in the beef industry. Unfortunately, this means of
appraising the beef
animal doesn't provide either the buyer or the seller with an evaluation of
the marbling of the meat
inside the beef animal. For example, one might raise ten beef calves that on
sale day each weigh
the same and visually (or actually measured) appear to be nearly identical in
conformation. Later,
the ten are sold to a feed lot and weigh the same as well as appear similar in
conformation. When
they are sold to be butchered, again, the ten weigh and appear identical but
the U.S.D.A. grading
finally comes into play. It is possible that the meat from one animal will be
rated PRIME, the meat
from three others rated CHOICE and the meat from the last six will be graded
SELECT. Any
combination of grading is possible after butchering, but at previous sales
there was no premium
paid for the potential PRIME animals) nor was anyone able to predict which
animal would be
PRIME, CHOICE, or SELECT.
There have been several means attempted to measure beef muscles and
intramuscular fat.
Some prior art systems use x-rays and/or CAT scans for measuring. These
methods have several
drawbacks. Often the animal cannot remain motionless for the duration of the
scan, which could
take several minutes. Also, the technicians) are required to wear protective
(i.e. lead-vest) x-ray
gear when x-ray scanning is used.
Other prior art systems have been developed using various types of Ultrasound
(and/or
sonogram) imaging. See, for example, U.S. Patent 5,398,290 entitled "System
for Measurement
of Intramuscular Fat in Cattle" of Brethour, issued Mar. 14, 1995. Even though
the scanning time
of real-time ultrasound is less than that of x-rays or CAT scans, these
systems still have limitations.
A liquid solution must be used between the ultrasound scanning device and the
animal's hide to
allow transmission of the sound waves into the animal's muscle that is
targeted for measuring.
Several hundred beef animals passing through an area (i.e. chute) used for
scanning can possibly
create a slippery mess of solution on the floor. Additionally, the resulting
ultrasound images may
need to be interpreted by a highly skilled technician and inaccuracies are
possible.
It is thus apparent that there is a need in the art for an improved method or
apparatus which
provides information about the size and marbling of the animal's muscles while
removing
CA 02389271 2002-04-24
WO 01/33493 PCT/US99/25360
3
requirements for a sound conducting liquid, and while reducing the skill
required to operate the
system. The present invention meets these and other needs in the art.
DISCLOSURE OF INVENTION
It is an aspect of the present invention to provide information about the
percentage of
intramuscular fat, called marbling, of meat inside a live animal or a carcass.
It is another aspect of the invention to provide information about the size of
muscles within
a live animal or a carcass.
Another aspect of the invention is to provide such marbling and size
information without
requiring interpretation by a skilled technician.
Still another aspect of the invention is to provide information about the
thickness of back
fat within a live animal or carcass.
A further aspect of the invention is to grade and classify animals after
analyzing the
intramuscular fat, back fat and muscle size within the animal.
The above and other aspects of the invention are accomplished in a
computerized system
that scans the animal using Magnetic Resonance Imaging (MRI). The scan data is
analyzed within
the computer to determine the marbling of the muscle scanned. The scan is
further analyzed to
determine the size of the muscle scanned.
The scan data is produced as digital pixel values within scan wave lines. The
pixel values
are coded as gray scale values wherein the gray scale value of each pixel is
representative of the
type of tissue scanned. The computer system thus classifies each pixel, based
upon its gray scale
value, as representing fat, cartilage, muscle or skeletal tissue. Once
classified, the percentage of
intramuscular fat can be calculated and presented on the screen of the
computer system.
Also, once the pixel data is classified, a perimeter around the muscle being
scanned can be
defined by separating the muscle tissue from the fat and skeletal tissue or
the surrounding muscle
sheath called a fascia. Once the perimeter of the muscle is defined, the area
of the muscle is
calculated and presented to the user of the system.
Multiple scans can be performed along one dimension of the animal so that the
volume of
the muscle can be calculated.
Once the intramuscular fat and muscle size and volume have been determined for
a
particular animal, the animal is graded and classified for ranking within like
kinds of animals.
CA 02389271 2002-04-24
WO 01/33493 PCT/US99/25360
4
DESCRIPTION OF THE DRAWINGS
The above and other aspects, features, and advantages of the invention will be
better
understood by reading the following more particular description of the
invention, presented in
conjunction with the following drawings, wherein:
Fig. 1 shows an animal being measured using the present invention;
Fig. 2 shows a cross section of a beef animal, illustrating an area for
measurement with the
present invention;
Fig. 3 shows a cross-section of a beef animal, and shows MRI Wave lines where
an MRI
scanner produces data of the texture of the muscle, fascia, fat and skeletal
structure
of the beef animal;
Fig. 4 illustrates pixels from an MRI wave line, and further illustrates which
pixels are
muscle and fat;
Fig. 5 illustrates how an area of a muscle is determined;
Fig. 6 shows a flowchart of the process of scanning an animal and calculating
a percentage
of marbling within the tissue of the animal;
Fig. 7 shows a flowchart of the process of determining a cross-sectional area
of a muscle
within an animal;
Fig. 8 shows a flowchart of the process of measuring a thickness of back fat
in a beef
animal;
Fig. 9 illustrates the process of Fig. 8; and
Fig. 10 shows a flowchart of the process of determining a volume of a muscle
within an
animal.
BEST MODE FOR CARRYING OUT THE INVENTION
The following description is of the best presently contemplated mode of
carrying out the
present invention. This description is not to be taken in a limiting sense but
is made merely for the
purpose of describing the general principles of the invention. The scope of
the invention should be
determined by referencing the appended claims.
Fig. 1 shows an animal being measured using the present invention. Referring
to Fig. l, a
chute 118 is used to contain an animal 102 being measured by the present
invention. Alternatively,
the animal could be standing at halter, or simply standing freely. A computer
system 104 is shown
CA 02389271 2002-04-24
WO 01/33493 PCT/US99/25360
having a display 106, keyboard 108 and mouse 110. This is a conventional
personal computer
system, which is commonly used. Cable 112 connects the computer system to the
MRI scanner
parts 114 and 116. Part 114 typically contains the electronics of the MRI
scanner, and part 116
contains the scanning element that is placed over the animal. An example of
this type of MRI
5 system can be found in U.S. Patent 5,304,930 entitled "Remotely Positioned
MRI System", issued
Apr. 19, 1994 to Crowley, et al.
The MRI scanning element 116 can be placed at any location over the animal
102, and
precisely located as desired. Also, scanning element 116 can be placed on a
movable apparatus (not
shown), controlled by the computer system 104, that allows the scanning
element 116 to be moved
along the length of the animal to obtain the multiple scans needed to perform
muscle volume
measurements, as described below. Once the scanning element 116 is in place,
the mouse 110, or
other switch device (not shown), is used to start a scan, which typically
takes less than one second.
Should the animal 102 move during the scan, the operator can re-scan so as to
get a correctly
focused scan.
Once the scan is complete, computer system 104 analyzes the MRI wave lines to
determine
the marbling percentage, size of the muscle, and thickness of back fat located
underneath the
scanning element 116, as will be described below. After analyzing these
traits, the computer system
104 grades and classifies the animal to rank it within like kinds of animals.
Once the animal is graded, the computer system 104 can direct the animal to
different
holding pens by opening a gate into the selected holding pen.
As an alternative to producing digital data directly from the MRI scanning
device, an MRI
scan can produce a photographic media print or negative, which is then scanned
and digitized using
a conventional scanner.
Fig. 2 shows a cross-sectional view, taken between the twelfth and thirteenth
ribs, of the
animal 102 (Fig. 1) and illustrates the location where the scan is typically
taken in a beef animal.
Referring to Fig. 2, circle 202 illustrates the location for a typical MRI
scan, which would produce
data showing the structure of the muscle fascia 203, the spine 204, cartilage
between the ribs 205,
muscle area 206, back fat 207, hide 208 and hair 209 of the animal 102.
Fig. 3 shows the area 202 and illustrates the MRI scan wave lines. Referring
to Fig. 3, the
MRI scan, created from the scanner element 116, produces a plurality of scan
wave lines 302 which
capture data about the fascia 203, spine 204, muscle 206, cartilage 205,
located between the ribs,
CA 02389271 2002-04-24
WO 01/33493 PCT/US99/25360
6
back fat 207, hide 208 and hair 209 of the animal being scanned. The wave
lines 302 extend
throughout the tissue area being scanned. A magnetic field of 0.5 to 1.5 tesla
can be used to
produce the scan wave lines 302. Producing one wave line takes approximately
one twentieth of
one second, so producing the eight wave lines shown take less than one half of
one second.
Additional scan lines could be produced, and the scan lines produced closer
together, to scan any
desired percentage of the muscle, up to 100 percent. Producing additional wave
lines takes
additional time, however, additional wave lines could be used to produce
additional accuracy in the
estimate of the marbling and size of the muscle. As more time is taken, to
produce a higher number
and/or density of wave lines, the probability that the animal will move
increases, but a complete
scan is possible if the animal does not move, which is possible for some
docile animals or by
restraining the any animal. In the preferred embodiment of the invention, less
than ten wave lines
are produced, requiring less than one half of one second. Also, when an MRI
scan is performed,
the distance between the scan wave lines is set within the MRI scanner, so the
depth of the tissue
scanned by ten wave lines is therefore also adjustable.
Fig. 4 shows one of the scan wave lines, and illustrates the pixel data that
is returned to the
computer system 104 as part of a scan wave line. Referring to Fig. 4, the scan
wave lines 302 are
shown as they were produced by the scan shown in Fig. 3. The area 402 is an
enlarged illustration
of the pixels that are part of one of the scan wave lines. Within the area
402, three rows of pixels,
labeled 404, 406, and 408, are shown. Three rows of pixels is by way of
example only, since the
number of pixels located within a scan wave line is variable, depending upon
the setting of the MRI
scanner. Additionally, many pixels can be combined into a cluster of pixels,
for example by
averaging gray scale values of all the pixels within each cluster, and the
cluster analyzed instead of
analyzing individual pixels.
Within the row 408, pixel 410 illustrates muscle tissue, and pixel 412
illustrates
intramuscular fat. Typically each of the pixels is returned as a gray scale
level, and as discussed
above several pixels may be combined before analysis to produce a combined
gray scale level. The
gray scale level for each pixel is analyzed to separate each into a pixel that
represents muscle tissue,
a pixel that represents fat tissue, a pixel that represents cartilage, or a
pixel that represents skeletal
tissue, wherein the separation is based upon the range of gray scale levels
typically found for each
of the type of tissues within the type of animal scanned. Also, a large
sequence of fat or skeletal
pixels could be ignored, if they are located on the periphery of the ribeye
muscle, assuming that they
CA 02389271 2002-04-24
WO 01/33493 PCT/US99/25360
7
represent a large fat area that would be trimmed from the meat, or they
represent a bone, that would
be trimmed from the meat. The remaining pixels are counted and the ratio of
intramuscular fat
pixels to muscle pixels is calculated and this ratio converted to a percentage
of marbling.
Fig. 5 shows a cross-section and illustrates calculating an area of the
muscle. Deferring to
Fig. 5, scan wave lines 302 are shown with an outline 502 around the ribeye
(longissimus dorsi)
muscle. Each scan wave line 302 is terminated by analyzing the pixels, as
illustrated in Fig. 4, by
terminating a line when a series of fascia, fat or skeletal pixels are found.
That is, the line is
examined, pixel by pixel, from the center of the line outward, and the end of
the line is set at the
beginning of a long series of fascia, fat, or skeletal pixels. The length of
the series of fascia, fat, or
skeletal pixels necessary to terminate a scan wave line is typically ten to
twenty pixels, although this
is variable depending upon the density of the scan, the type of muscle and the
type of animal. Once
the termination points of each scan wave line are fixed, the length of the
scan wave line is
calculated as the distance between the termination points.
The outline 502 is created by connecting the ends of the scan wave lines 302.
The line 504
is located by using the center of the uppermost scan wave line and extending
upward a distance 510
equal to one-half the distance between scan wave lines, then forming two right
triangles 506 and
508.
A line 512 is extended from each end of each scan wave line to the next scan
wave line
below. The two scan wave lines and the two extended lines thus form a
rectangle. A right triangle
is formed at the end of each rectangle by connecting the extended line, the
remaining part of the
longer scan wave line and the ends of the two scan wave lines. The area
between the two lines is
calculated by calculating the area of the rectangle and the two right
triangles. After the area
between all the scan lines is calculated, the areas, including right triangles
506 and 508 are summed
to get the total area of the ribeye muscle.
If the volume of the muscle is desired, multiple MRI scans are made, using a
fixed distance
between the scans. By calculating the area of the muscle at each scan, and
calculating the volume
between each scan in the same manner as the area is calculated, the volume of
the muscle can be
calculated.
In a similar manner, the thickness of the fat area between the muscle and the
hide of the
animal can be calculated as described below with respect to Figs. 8 and 9.
Because the hide of a beef animal is denser than the fat or muscle, and
differs in density
CA 02389271 2002-04-24
WO 01/33493 PCT/US99/25360
8
from the skeletal structure, the thickness of the hide can also be determined
using these methods.
Fig. 6 shows a flowchart of the process of the invention for determining the
marbling and
area of the muscle. This flowchart is called within scanner software contained
in the computer
system 104 (Fig. 1) when the operator requests a scan after the scanner has
been positioned over
the desired part of the animal. Referring to Fig. 6, after entry, step 602
sends a command to the
scanner to perform a scan. After the scan is complete, and the scan data is
returned to the computer
system 104, step 604 classifies all the pixels found, as described above, and
step 606 terminates the
scan lines by removing the pixels that represent fat, cartilage or skeletal
tissue at the ends of the
scan lines, as described above with respect to Fig. 5. Step 608 gets the first
pixel of the first scan
wave line from the scan data. Step 610 adds this pixel to the pixel count.
Step 612 determines
whether the pixel represents fat, and if it does, step 612 goes to step 614
which increments the
intramuscular fat count of pixels. Step 616 determines if there are more
pixels to retrieve, and if
so, step 616 goes back to step 606 to process the next pixel.
After all pixels have been processed, step 618 calculates the percent of
marbling by dividing
the intramuscular fat count of pixels by the total count of pixels, and
multiplying the result by 100
to get the percentage. Step 620 then calls Fig. 7 to calculate the area of the
muscle, step 622 calls
Fig. 8 to get the back fat thickness, and step 624 calls Fig. 10 to get the
muscle volume.
After these have been calculated, step 626 displays the marbling percentage,
muscle area,
back fat thickness, and muscle volume on the display 106 (Fig. 1 ).
Fig. 7 shows a flowchart of the process of calculating the area of the muscle,
as described
above with respect to Fig. 5. Referring to Fig. 7, after entry, step 702 gets
the first wave line, which
has already been processed, as described above with respect to step 604 (Fig.
6) and Fig. 5. Step
704 gets the next (second) wave line after the first wave line, and step 706
calculates the area of the
rectangle between the two wave lines. Step 708 calculates the area of the two
triangles at each end
of the lines, and step 710 adds the area of the rectangle and the two
triangles to the total area.
Step 712 determines if there are more wave lines in the scan, and if so, goes
to block 714,
which copies the second wave line to the first wave line, so that it can be
used in the next
calculation. Step 714 then returns to step 704 to process the next wave line.
After all wave lines have been processed, step 712 goes to step 716 which
calculates the area
of the two triangles at the top of the muscle, and step 718 returns the muscle
area to Fig. 6, where
it is displayed.
CA 02389271 2002-04-24
WO 01/33493 PCT/US99/25360
9
Although the invention has been described as measuring the intramuscular fat,
area and
volume of muscles within live animals, those skilled in the art will recognize
that the invention can
also be used to achieve these same measurements within the carcass of a
butchered animal.
Fig. 8 shows a flowchart of calculating the back fat thickness in a beef
animal, as is called
from step 622 of Fig. 6. Referring to Fig. 8, after entry, step 802 determines
the outer perimeter of
the muscle, as described above with respect to Fig. 5. Step 804 determines the
outer perimeter of
the back fat in the same manner described above to find the perimeter of the
muscle. Step 806
locates the left and right intersection points of the top most scan wave line
with the muscle
perimeter. Step 808 locates the left and right intersection points of the next
to the top most scan
wave line and the muscle perimeter. Step 810 determines the distance between
the two left points
and the distance between the two right points and selects the set of left or
right points having the
greatest distance between them.
Step 812 constructs a line between the two points selected in step 810 and
step 814 locates
the center of the line constructed in step 812. Step 816 constructs a line
perpendicular to the line
constructed in step 812 at the center located in step 814 and in a direction
toward the top scan wave
line. Step 818 locates the intersection of the line constructed in step 816
and the muscle perimeter,
and step 820 locates the intersection of the line constructed in step 816 and
the back fat perimeter.
Step 822 determines the distance between the intersection located in step 818
and the intersection
located in step 820 and step 824 returns this distance as the back fat
thickness to Fig. 6.
Fig. 9 graphically depicts the process of Fig. 8 of finding the back fat
thickness. Referring
to Fig. 9, wave scan line 902 is the top most wave scan line that intersects
the muscle perimeter, and
wave scan line 904 is the next to the top most wave scan line that intersects
the muscle perimeter.
Points 906 and 910 are the respective left and right intersection points of
the top most wave scan
line with the muscle perimeter. Points 908 and 912 are the respective left and
right intersection
points of the next to the top most wave scan line with the muscle perimeter.
Since left points 906
and 908 are farther apart than right points 910 and 912, points 906 and 908
would be selected and
a line constructed between them. The center of this line is located and a line
914 is drawn
perpendicular to the line between points 906 and 908 at the center point, in a
direction toward the
top most scan wave line 902. The intersection 916 with this line and the
muscle perimeter and the
intersection 918 with this line and the back fat perimeter are located and the
distance between them
is the back fat thickness.
CA 02389271 2002-04-24
WO 01/33493 PCT/US99/25360
Fig. 10 shows a flowchart of calculating the volume of a muscle, as called
from step 624 of
Fig. 6. Referring to Fig. 10, after entry, step 1002 sends another command to
the MRI scanner to
scan another image at a known distance from the first image scanned in Fig. 6.
Step 1004 classifies
the pixels of the scanned data, step 1006 terminates the scan lines from the
second image scanned
5 in step 1002, and step 1008 connects the ends of corresponding scan lines
from the scan performed
in Fig. 6, and the scan performed in step 1002. Step 1010 calls Fig. 7 to
calculate the area of the
scan performed in step 1002, and step 1012 then calculates the volume between
the two scan images
in the same manner the areas of the images were calculated.
Step 1014 adds this volume to the total volume accumulated, and block 1016
determines
10 if more images need to be scanned. This determination is based on the size
of the image scanned
in step 1002, and terminates when the area of the scan becomes small enough to
be the end of the
muscle. If more images are needed, step 1016 returns to step 1002 to scan the
next image.
After all images are scanned, and volumes calculated, step 1016 returns the
volume to Fig.
6 for display.
Appendix 1, attached hereto and incorporated herein by reference for all that
is disclosed
and taught therein, shows an example of an MRI scan of a carcass and
illustrates determining the
size and marbling of the tissue scanned.
Having described a presently preferred embodiment of the present invention, it
will be
understood by those skilled in the art that many changes in construction and
circuitry and widely
differing embodiments and applications of the invention will suggest
themselves without departing
from the scope of the present invention, as defined in the claims. The
disclosures and the
description herein are intended to be illustrative and are not in any sense
limiting of the invention,
defined in scope by the following claims.
