U.S. patent application number 14/984679 was filed with the patent office on 2016-07-28 for grayscale thermographic imaging.
The applicant listed for this patent is Kadambari Nuguru, James G. Spahn. Invention is credited to Kadambari Nuguru, James G. Spahn.
Application Number | 20160213304 14/984679 |
Document ID | / |
Family ID | 56082951 |
Filed Date | 2016-07-28 |
United States Patent
Application |
20160213304 |
Kind Code |
A1 |
Spahn; James G. ; et
al. |
July 28, 2016 |
Grayscale Thermographic Imaging
Abstract
Through the measurement and interpretation of the pixels of
grayscale digital thermographic images of abnormalities of the skin
and its subcutaneous tissue, early intervention and treatment of
abnormalities of the skin and its subcutaneous tissue are possible,
thereby assisting clinicians in making significant impacts on
prevention and treatment.
Inventors: |
Spahn; James G.; (Carmel,
IN) ; Nuguru; Kadambari; (Indianapolis, IN) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Spahn; James G.
Nuguru; Kadambari |
Carmel
Indianapolis |
IN
IN |
US
US |
|
|
Family ID: |
56082951 |
Appl. No.: |
14/984679 |
Filed: |
December 30, 2015 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13439177 |
Apr 4, 2012 |
9357963 |
|
|
14984679 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
A61B 5/015 20130101;
A61B 5/445 20130101; G06T 2207/20072 20130101; A61B 5/0082
20130101; G06T 7/0016 20130101; A61B 2576/02 20130101; G06T
2207/30088 20130101; A61B 5/1075 20130101; G06T 7/11 20170101; A61B
5/6844 20130101; A61B 5/0077 20130101; G06T 2207/10048 20130101;
A61B 5/1072 20130101 |
International
Class: |
A61B 5/00 20060101
A61B005/00; G06T 7/00 20060101 G06T007/00 |
Claims
1. A method of detecting abnormalities in mammalian skin and
subcutaneous tissue, said method comprising the steps of: providing
a infrared imaging device, said device comprising a first laser
emitting device and a second laser emitting device, wherein beams
from said first and second laser emitting devices converge at a
predetermined distance from a target; arranging said imaging device
at a distance from said target such that said beams converge;
defining a first skin area on a mammal; acquiring a long-wave
infrared image of said first skin area from said predetermined
distance using said imaging device; defining a pixel value for a
predetermined range of temperature in said first skin area;
calculating an average pixel value for said first skin area;
defining a second skin area on a mammal; acquiring a long-wave
infrared image of said second skin area from said predetermined
distance using said imaging device; defining a pixel value for a
predetermined range of temperature in said second skin area;
calculating an average pixel value for said second skin area;
calculating an average pixel value for said second skin area; and
calculating a ratio between said first skin area average pixel
value and said second skin area average pixel value.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] The present divisional application claims priority to
Non-Provisional patent application Ser. No. 13/439,177, filed Apr.
4, 2012.
BACKGROUND
[0002] Over the last century, clinicians, which term includes
herein certified and licensed medical doctors of all specialties,
osteopathic doctors of all specialties, podiatrists, dental doctors
of all specialties, chiropractors, veterinarians of all
specialties, nurses, and medical imaging technicians, have become
dependent on the use of medical devices that assist them in their
delivery of patient-centered care. The common function of these
devices is to assist and not replace the clinical judgment of the
clinician. This fulfills the dictum that best practice is clinical
judgment assisted by scientific data and information.
[0003] Entering into the era of computer science and sophisticated
electronics, clinicians have the opportunity to be supported by
data and information in a statistically significant and timely
fashion. These advancements have allowed more extensive and useful
collection of meaningful data that can be acquired, analyzed, and
applied in conjunction with the knowledge and expertise of the
clinician.
[0004] Medical long-wave infrared (LIR) thermography has been known
to be beneficial in the evaluation of thermal heat intensity and
gradiency relating to abnormalities of the skin and subcutaneous
tissue (SST). Although this technology has expanded to other areas
of medical evaluation, the scope of this patent application is
limited to the SST abnormalities. These abnormalities include the
formation of deep tissue injury (DTI) and subsequent necrosis
caused by mechanical stress, infection, auto-immune condition, and
vascular flow problems. DTI caused by mechanical stress (pressure,
shear and frictional forces) can be separated into three
categories. The first category is a high magnitude/short duration
mechanical stress represented by traumatic and surgical wounds. The
second category is low magnitude/long duration mechanical stress
represented by pressure ulcer development, which is also a factor
in the development of ischemic and neuropathic wounds. The third
category is a combination of categories one and two represented by
pressure ulcer formation in the bariatric patient.
[0005] The pathophysiologic conditions that occur with DTI and
subsequent necrosis of the affected tissue are ischemia, cell
distortion, impaired lymphatic drainage, impaired interstitial
fluid flow, and reperfusion injury: Category one is dominated by
cell distortion and even destruction. Category two is dominated by
ischemia. Category three is a combination of cell distortion and
ischemia.
[0006] Hypoxia causes aerobic metabolism to convert to anaerobic
metabolism. This occurrence causes lactic acidosis followed by cell
destruction, release of enzymes and lytic reactions. The release of
these substances causes additional cell injury and destruction, and
initiation of the inflammatory response.
[0007] It is very important to recognize that ischemic-reperfusion
injury is associated with all of the above mechanical stress
induced SST injuries. This condition is caused by a hypoxia induced
enzymatic change and the respiratory burst associated with
phagocytosis when oxygen returns after an ischemic event. The
result of ischemic-reperfusion injury is the formation of oxygen
free radicals (hydroxyl, superoxide, and hydrogen peroxide) that
cause damage to healthy and already injured cells leading to
extension of the original injury.
[0008] SST injury and subsequent necrosis can also be caused by
vascular disorders. Hypoxia can be caused by an arterial occlusion
or by venous hypertension. Lymphatic flow or node obstruction can
also create vascular induced injury by creating fibrous restriction
to venous drainage and subsequent cellular stasis in the capillary
system. These disorders are also accentuated by reperfusion injury
and oxygen free radical formation.
[0009] Infection of the skin (impetigo), subcutaneous tissue
(cellulitis), deep tissue (fasciitis), bone (osteomyelitis) and
cartilage (chondritis) causes injury and necrosis of the affected
tissue. Cells can be injured or destroyed by the microorganism
directly, by toxins released by the microorganism and/or the
subsequent immune and inflammatory response. These disorders are
also accentuated by reperfusion injury and oxygen free radical
formation.
[0010] Auto-immune morbidities of the skeletal joints (rheumatoid
arthritis), subcutaneous tissue (tendonitis, myelitis, dermatitis)
and blood vessels (vasculitis) cause similar dysfunction and
necrosis of the tissue being affected by the hypersensitivity
reactions on the targeted cells and the subsequent inflammatory
response. Again, these conditions are accentuated by reperfusion
and oxygen free radical formation.
[0011] The common event that addresses all of the above SST
injuries is the inflammatory response. This response has two
stages. The first stage is vascular and the second is cellular. The
initial vascular response is vasoconstriction that will last a
short time. The constriction causes decrease blood flow to the area
of injury. The decrease in blood flow causes vascular "pooling" of
blood (passive congestion) in the proximal arterial vasculature in
the region of injury and intravascular cellular stasis occurs along
with coagulation.
[0012] The second vascular response is extensive vasodilation of
the blood vessels in the area of necrosis. This dilation along with
the "pooled" proximal blood causes increased blood flow with high
perfusion pressure into the area of injury. This high pressure flow
can cause damage to endothelial cells. Leakage of plasma, protein,
and intravascular cells causes more cellular stasis in the
capillaries (micro-thrombotic event) and hemorrhage into the area
of injury. When the perivascular collagen is injured, intravascular
and extravascular coagulation occurs. The rupture of the mast cells
causes release of histamine that increases the vascular dilation
and the size of the junctions between the endothelial cells. This
is the beginning of the cellular phase. More serum and cells
(mainly neutrophils) enter into the area of the mixture of injured
and destroyed cells by the mechanism of marginalization, emigration
(diapedesis) and the chemotaxic recruitment (chemotaxic gradiency).
Stalling of the inflammatory stage can cause the area of necrosis
(ring of ischemia) to remain in the inflammatory stage long past
the anticipated time of 2-4 days. This continuation of the
inflammatory stage leads to delayed resolution of the ischemic
necrotic event.
[0013] The proliferation stage starts before the inflammatory stage
recedes. In this stage angiogenesis occurs along with formation of
granulation and collagen deposition. Contraction occurs, and peaks,
at 5-15 days post injury.
[0014] Re-epithelialization occurs by various processes depending
on the depth of injury. Partial thickness wounds can resurface
within a few days. Full thickness wounds need granulation tissue to
form the base for re-epithelialization to occur. The full thickness
wound does not heal by regeneration due to the need for scar tissue
to repair the wound. The repaired scarred wound has less
vascularity and tensile strength of normal regional uninjured SST.
The final stage is remodeling. In this stage the collagen changes
from type III to a stronger type I and is rearranged into an
organized tissue.
[0015] All stages of wound healing require adequate vascularization
to prevent ischemia, deliver nutrients, and remove metabolic waste.
Following the vascular flow and metabolic activity of a necrotic
area is currently monitored by patient assessment and clinical
findings of swelling, pain, redness, increased temperature, and
loss of function.
SUMMARY
[0016] Having a real time control allows an area of interest (AOI)
to be recognized. The AOI can be of greater intensity (hotter) or
less intensity (cooler) than the normal SST of that region of the
body. The AOI can then be evaluated by the clinician for the degree
of metabolism, blood flow, necrosis, inflammation and the presence
of infection by comparing the warmer or cooler thermal intensity of
the AOI or wound base and peri-AOI or wound area to the normal SST
of the location being imaged. Serial imaging also can assist the
clinician in the ability to recognize improvement or regression of
the AOI or wound over time.
[0017] The use of an LIR thermal and digital visual imager can be a
useful adjunct tool for clinicians with appropriate training to be
able to recognize physiologic and anatomical changes in an AOI
before it presents clinically and also the status of the AOI/wound
in a trending format. By combining the knowledge obtained from the
images with a comprehensive assessment, skin and subcutaneous
tissue evaluation, and an AOI or wound evaluation will assist the
clinician in analyzing the etiology, improvement or deterioration,
and the presence of infection affecting the AOI or wound.
[0018] The foundational scientific principles behind LIR
thermography technology are energy, heat, temperature, and
metabolism.
[0019] Energy is not a stand-alone concept. Energy can be passed
from one system to another, and can change from one form to
another, but can never be lost. This is the First Law of
Thermodynamics. Energy is an attribute of matter and
electromagnetic radiation. It is observed and/or measured only
indirectly through effects on matter that acquires, loses or
possesses it and it comes in many forms such as mechanical,
chemical, electrical, radiation (light), and thermal.
[0020] The present application focuses on thermal and chemical
energy. Thermal energy is the sum of all of the microscopic scale
randomized kinetic energy within a body, which is mostly kinetic
energy. Chemical energy is the energy of electrons in the force
field created by two or more nuclei; mostly potential energy.
[0021] Energy is transferred by the process of heat. Heat is a
process in which thermal energy enters or leaves a body as the
result of a temperature difference. Heat is therefore the transfer
of energy due to a difference in temperature; heat is a process and
only exists when it is flowing. When there is a temperature
difference between two objects or two areas within the same object,
heat transfer occurs. Heat energy transfers from the warmer areas
to the cooler areas until thermal equilibrium is reached. This is
the Second Law of Thermodynamics. There are four modes of heat
transfer: evaporation, radiation, conduction and convection.
[0022] Molecules are the workhorses and are both vehicles for
storing and transporting energy and the means of converting it from
one form to another. When the formation, breaking, or rearrangement
of the chemical bonds within the molecules is accompanied by the
uptake or release of energy it is usually in the form of heat. Work
is completely convertible to heat and defined as a transfer due to
a difference in temperature, however work is the transfer of energy
by any process other than heat. In other words, performance of work
involves a transformation of energy.
[0023] Temperature measures the average randomized motion of
molecules (kinetic energy) in a body. Temperature is an intensive
property by which thermal energy manifests itself. It is measured
by observing its effect on some temperature dependent variable on
matter (i.e. ice/steam points of water). Scales are needed to
express temperature numerically and are marked off in uniform
increments (degrees).
[0024] As a body loses or gains heat, its temperature changes in
direct proportion to the amount of thermal energy transferred from
a high temperature object to a lower temperature object. Skin
temperature rises and falls with the temperature of the
surroundings. This is the temperature that is referred to in
reference to the skins ability to lose heat its surroundings.
[0025] The temperature of the deep tissues of the body (core
temperatures) remains constant (within .+-.1.degree.
F./.+-.0.6.degree. C.) unless the person develops a febrile
illness. No single temperature can be considered normal.
Temperature measurements on people who had no illness have shown a
range of normal temperatures. The average core temperature is
generally considered to be between 98.0.degree. F. and 98.6.degree.
F. measured orally or 99.0.degree. F. and 99.6.degree. F. measured
rectally. The body can temporarily tolerate a temperature as high
as 101.degree. F. to 104.degree. F. (38.6.degree. C. to 40.degree.
C.) and as low as 96.degree. F. (35.5.degree. C.) or lower.
[0026] Metabolism simply means all of the chemical reactions in all
of the cells of the body. Metabolism creates thermal energy. The
metabolic rate is expressed in terms to the rate of heat release
during the chemical reactions. Essentially all the energy expended
by the body is eventually converted into heat.
[0027] Since heat flows from hot to cold temperature and the body
needs to maintain a core temperature of 37.0.degree.
C..+-.0.75.degree. C., the heat is conserved or dissipated to the
surroundings. The core heat is moved to the skin surface by blood
flow. Decreased flow to the skin surface helps conserve heat, while
increased flow promotes dissipation. Conduction of the core heat to
the skin surface is fast, but inadequate alone to maintain the core
temperature. Heat dissipation from the skin surface (3 mm
microclimate) also occurs due to the conduction, convection and
evaporation.
[0028] Heat production is the principal by-product of metabolism.
The rate of heat production is called the metabolic rate of the
body. The important factors that affect the metabolic rate are:
[0029] 1. Basal Rate of Metabolism (ROM) of all cells of the
body.
[0030] 2. Extra ROM caused by muscle activity including
shivering.
[0031] 3. Extra ROM caused by the effect of thyroxine and other
hormones to a less extent (i.e.: growth hormone, testosterone).
[0032] 4. Extra ROM caused by the effect of epinephrine,
norepinephrine, and sympathetic stimulation on the cells.
[0033] 5. Extra ROM caused by increased chemical activity in the
cells themselves, especially when the cell temperature
increases.
[0034] Most of the heat produced in the body is generated in the
deep organs (liver, brain, heart and the skeletal muscles during
exercise). The heat is then transferred to the skin where the heat
is lost to the air and other structures. The rate that heat is lost
is determined by how fast heat can be conducted from where it is
produced in the body core to the skin.
[0035] The skin, subcutaneous tissues and especially adipose tissue
are the heat insulators for the body. The adipose tissue is
important since it conducts heat only 33% as effective as other
tissue and specifically 52% as effective as muscle. Conduction rate
of heat in human tissue is 18 kcal/cm/m2k. The subcutaneous tissue
insulator system allows the core temperature to be maintained yet
allowing the temperature of the skin to approach the temperature of
the surroundings.
[0036] Blood flows to the skin from the body core in the following
manner. Blood vessels penetrate the adipose tissue and enter a
vascular network immediately below the skin. This is where the
venous plexus comes into play. The venous plexus is especially
important because it is supplied by inflow from the skin
capillaries and in certain exposed areas of the body
(hands-feet-ears) by the highly muscular arterio-venous
anastomosis. Blood flow can vary in the venous plexus from barely
above zero to 30% of the total cardiac output. There is an
approximate eightfold increase in heat conductance between the
fully vasoconstricted state and the fully vasodilated state. The
skin is an effective controlled heat radiator system and the
controlled flow of blood to the skin is the body's most effective
mechanism of heat transfer from the core to the surface.
[0037] Heat exchange is based on the scientific principle that heat
flows from warmer to cooler temperatures. Temperature is thought of
as heat intensity of an object. The methods of heat exchange are:
radiation (60%), loss of heat in the form of LIR waves (thermal
energy), conduction to a solid object (3%), transfer of heat
between objects in direct contact and loss of heat by conduction to
air (15%) caused by the transfer of heat, caused by the kinetic
energy of molecular motion. Much of this motion can be transferred
to the air if it is cooler than the surface. This process is
self-limited unless the air moves away from the body. If that
happens, there is a loss of heat by convection. Convection is
caused by air currents. A small amount of convection always occurs
due to warmer air rising. The process of convection is enhanced by
any process that moves air more rapidly across the body surface
(forced convection). This includes fans, air flow beds and air
warming blankets.
[0038] Convection can also be caused by a loss of heat by
evaporation which is a necessary mechanism at very high air
temperatures. Heat (thermal energy) can be lost by radiation and
conduction to the surroundings as long as the skin is hotter than
the surroundings. When the surrounding temperature is higher than
the skin temperature, the body gains heat by both radiation and
conduction. Under these hot surrounding conditions the only way the
body can release heat is by evaporation. Evaporation occurs when
the water molecule absorbs enough heat to change to gas. Due to the
fact water molecules absorb a large amount of heat in order to
change into a gas, large amounts of body heat can be removed from
the body.
[0039] Insensible heat loss dissipates the body's heat and is not
subject to body temperature control (water loss through the lungs,
mouth and skin). This accounts for 10% heat loss produced by the
body's basal heat production. Sensible heat loss by evaporation
occurs when the body temperature rises and sweating occurs.
Sweating increases the amount of water to the skins surface for
vaporization. Sensible heat loss can exceed insensible heat loss by
30 times. The sweating is caused by electrical or excess heat
stimulation of the anterior hypothalamus pre optic area.
[0040] The role of the hypothalamus (anterior pre-optic area) in
the regulation of the body's temperatures occurs due to nervous
feedback mechanisms that determine when the body temperature is
either too hot or too cold.
[0041] The role of temperature receptors in the skin and deep body
tissues relate to cold and warm sensors in the skin. Cold sensors
outnumber warm sensors 10 to 1. The deep tissue receptors occur
mainly in the spinal cord, abdominal viscera and both in and around
the great veins. The deep receptors mainly detect cold rather than
warmth. These receptors function to prevent low body temperature.
These receptors contribute to body thermoregulation through the
bilateral posterior hypothalamus area. This is where the signals
from the pre-optic area and the skin and deep tissue sensors are
combined to control the heat producing and heat conserving
reactions of the body.
[0042] Temperature Decreasing Mechanisms:
[0043] 1. Vasodilation of all blood vessels, but with intense
dilation of skin blood vessels that can increase the rate of heat
transfer to the skin eight fold.
[0044] 2. Sweating can remove 10 times the basal rate of body heat
with an additional 1.degree. C. increase in body temperature.
[0045] 3. Decrease in heat production by inhibiting shivering and
chemical thermogenesis.
[0046] Temperature Increasing Mechanisms:
[0047] 1. Skin vasoconstriction throughout the body.
[0048] 2. Increase in heat production by increasing metabolic
activity. [0049] a. Shivering [0050] i. 4 to 5 times increase
[0051] b. Chemical Thermogenesis (brown fat) [0052] i. Adults
10-15% increase [0053] ii. Infants 100% increase
[0054] LIR thermography evaluates the infra-red thermal intensity.
The microbolometer is a 320.times.240 pixel array sensor that can
acquire the long-wave infrared wavelength (7-14 micron) (NOT
near-infrared thermography) and convert the thermal intensity into
electrical resistance. The resistance is measured and processed
into digital values between 1-254. A digital value represents the
long-wave infrared thermal intensity for each of the 76,800 pixels.
A grayscale tone is then assigned to the 1-254 thermal intensity
digital values. This allows a grayscale image to be developed.
[0055] Using LIR thermography is a beneficial device to monitor
metabolism and blood flow in a non-invasive test that can be
performed bedside with minimal patient and ambient surrounding
preparation. The ability to accurately measure the LIR thermal
intensity of the human body is made possible because of the skins
emissivity (0.98.+-. is 0.01), which is independent of
pigmentation, absorptivity (0.98.+-.0.01) reflectivity (0.02) and
transmitability (0.000). The human skin mimics the "BlackBody"
radiation concept. A perfect blackbody only exists in theory and is
an object that absorbs and reemits all of its energy. Human skin is
nearly a perfect blackbody as it has an emissivity of 0.98,
regardless of actual skin color. These same properties allow
temperature degrees to be assigned to the pixel digital value. This
is accomplished by calibration utilizing a "BlackBody" simulator
and an algorithm to account for the above factors plus ambient
temperatures. A multi-color palate can be developed by clustering
pixel values. There are no industry standards how this should be
done so many color presentations are being used by various
manufacturers. The use of gray tone values is standardized,
consistent and reproducible. Black is considered cold and white is
considered hot by the industry.
[0056] An LIR camera has the ability to detect and display the LIR
wavelength in the electromagnetic spectrum. The basis for infrared
imaging technology is that any object whose temperature is above
0.degree. K radiates infrared energy. Even very cold objects
radiate some infrared energy. Even though the object might be
absorbing thermal energy to warm itself, it will still emit some
infrared energy that is detectable by sensors. The amount of
radiated energy is a function of the object's temperature and its
relative efficiency of thermal radiation, known as emissivity.
[0057] Emissivity is a measure of a surface's efficiency in
transferring infrared energy. It is the ratio of thermal energy
emitted by a surface to the energy emitted by a perfect blackbody
at the same temperature.
[0058] LIR thermography is a beneficial device to monitor
metabolism and blood flow in a non invasive test that can be
performed bedside with minimal patient and ambient surrounding
preparation. It uses the scientific principles of energy, heat,
temperature and metabolism. Through measurement and interpretation
of thermal energy, it produces images that will assist clinicians
to make a significant impact on wound care (prevention, early
intervention and treatment) through detection.
[0059] In the method of grayscale digital thermographic imaging of
abnormalities of the skin and its subcutaneous tissue, the
improvement comprising: means for increasing and decreasing pixel
value brightness by adding a positive or negative offset to the raw
pixel value.
[0060] One embodiment of the present invention is in the method of
grayscale digital thermographic imaging of abnormalities of the
skin and its subcutaneous tissues, the improvement comprising
methods and apparatus for defining pixel intensity variations of a
long wave infrared image by measuring the thermal intensity ratio
of the average of all pixel values from a skin abnormality region
to the average of all pixel values from unaffected skin
regions.
[0061] Another embodiment of the present invention is in the method
of grayscale digital thermographic imaging of abnormalities of the
skin and its subcutaneous tissues, the improvement comprising
methods and apparatus for maintaining the separation of a
thermographic imager from skin at a set distance by converging two
light beams emanating from the imager at a point that is the set
distance for the imager to be from skin.
[0062] Another embodiment of the present invention is in the method
of grayscale digital thermographic imaging of abnormalities of the
skin at its subcutaneous tissues, the improvement comprising
methods and apparatus for obtaining the linear length and width
measurements of abnormalities and their square area.
[0063] Another embodiment of the present invention is in the method
of grayscale digital thermographic imaging of the skin and its
subcutaneous tissues, the improvement comprising methods and
apparatus for highlighting the digital thermographic image of an
area of skin to be measured and calculating the area of the
highlighted portion of the image in square centimeters by
determining the total number of pixels highlighted.
[0064] Another embodiment of the present invention is in the method
of grayscale digital thermographic imaging of the skin and its
subcutaneous tissues, the improvement comprising methods and
apparatus for encircling an area of interest and generating a
histogram of the encircled area to project the distribution of
pixel values therein.
[0065] Another embodiment of the present invention is in the method
of grayscale digital thermographic imaging of the skin and its
subcutaneous tissues, the improvement comprising methods and
apparatus for plotting profile lines in or through an area of skin
that is of interest and comparing it with a corresponding profile
line of normal skin.
BRIEF DESCRIPTION OF THE DRAWINGS
[0066] The present invention will be understood more fully from the
detailed description given hereinafter and from the accompanying
drawings of the preferred embodiment of the present invention,
which, however, should not be taken to limit the invention, but are
for explanation and understanding only.
[0067] In the drawings:
[0068] FIG. 1 shows a medical long wave infrared (LIR) and visual
views compared.
[0069] FIG. 2 shows a thermal span with default configuration
settings.
[0070] FIG. 3 shows an effect of adding a positive offset of the
thermal span.
[0071] FIG. 4 shows effect of adding negative offset on the thermal
span.
[0072] FIG. 5 shows a thermal image of a hand taken with default
settings.
[0073] FIG. 6 shows a thermal image of the hand when a positive
offset is added.
[0074] FIG. 7 shows a normal and abnormal selections made from a
thermal image and the corresponding results.
[0075] FIG. 8 shows an original image (left side) and thermal image
(right side--zoomed in) with abnormal selections made.
[0076] FIG. 9 shows a schematic representing pixel intensity
recognition (zoomed).
[0077] FIG. 10 shows a diagram of laser lights implementation.
[0078] FIG. 11 shows an experimental setup used to determine
digital camera and long-wave infrared microbolometer angels of
inclination.
[0079] FIG. 12 shows an embodiment of laser lights at an 18 inch
distance.
[0080] FIG. 13 shows a length and width measurements form an area
of interest.
[0081] FIG. 14 shows a schematic representing pixel intensity
recognition (zoomed.)
[0082] FIG. 15 shows a periwound region including the wound base
highlighted as area of interest and the results obtained for the
area selected.
[0083] FIG. 16 shows an area including normal, periwound and the
wound base regions highlighted as area of interest and the
corresponding results obtained for the area selected.
[0084] FIG. 17 shows wound histograms.
[0085] FIG. 18 shows normal histograms.
[0086] FIG. 19 shows a profile line showing the variation in the
grayscale values along the line drawn over an area of interest.
[0087] FIG. 20 shows comparing the Profile Line with the Reference
Line.
[0088] FIG. 21 shows a thermal Profile Line.
[0089] FIG. 22 shows a figure illustrating the formula for
calculating area under the curve.
[0090] FIG. 23 shows calculating areas above and below the selected
normal.
[0091] FIG. 24 shows a profile Line drawn through three
fingers.
[0092] FIG. 25 shows a profile Line plot on a graph.
[0093] Corresponding reference characters indicate corresponding
parts throughout the several views. The exemplary embodiments set
forth herein are not to be construed as limiting the scope of the
invention in any manner.
DETAILED DESCRIPTION OF THE EMBODIMENTS
[0094] The present invention will be discussed hereinafter in
detail in terms of various exemplary embodiments according to the
present invention with reference to the accompanying drawings. In
the following detailed description, numerous specific details are
set forth in order to provide a thorough understanding of the
present invention. It will be obvious, however, to those skilled in
the art that the present invention may be practiced without these
specific details. In other instances, well-known structures are not
shown in detail in order to avoid unnecessary obscuring of the
present invention.
[0095] Thus, all of the implementations described below are
exemplary implementations provided to enable persons skilled in the
art to make or use the embodiments of the disclosure and are not
intended to limit the scope of the disclosure, which is defined by
the claims. As used herein, the word "exemplary" or "illustrative"
means "serving as an example, instance, or illustration." Any
implementation described herein as "exemplary" or "illustrative" is
not necessarily to be construed as preferred or advantageous over
other implementations. Moreover, in the present description, the
terms "upper", "lower", "left", "rear", "right", "front",
"vertical", "horizontal", and derivatives thereof shall relate to
the invention as oriented in FIG. 1.
[0096] Furthermore, there is no intention to be bound by any
expressed or implied theory presented in the preceding technical
field, background, brief summary or the following detailed
description. It is also to be understood that the specific devices
and processes illustrated in the attached drawings, and described
in the following specification, are simply exemplary embodiments of
the inventive concepts defined in the appended claims. Hence,
specific dimensions and other physical characteristics relating to
the embodiments disclosed herein are not to be considered as
limiting, unless the claims expressly state otherwise.
[0097] Thermal images taken of the skin surface are constructed by
passively reading emitted radiant energy formed by the subcutaneous
tissue and the skin tissue by detecting wavelengths in the
long-wave infrared range (LIR) of 7-14 microns, and then in real
time converting these values into pixels within a digital image.
The value assigned to the pixel indicates the thermal intensities
of a particular area of the skin when imaged. The thermal images in
this embodiment are presented in digital unsigned (not having a
plus or minus sign) 8-bit grayscale with pixel values ranging from
0-254, however these same techniques work with images of varying
color resolutions. These images could be stored in the data bank
along with the information about the data the image has captured so
that it can be retrieved by a clinician for future review and
analysis. Generally, the unaffected skin thermal intensity will be
a uniform gray color within a range of +/-3 to 6 pixel values,
which is equal to 0.25 to 0.5 degrees centigrade. Abnormally hot
areas of the skin will be represented by patches of increasingly
white pixels, while abnormally cold areas will be represented by
increasingly dark patches of pixels.
[0098] The use of LIR (7-14 microns) imaging along with visual
digital imaging allows both physiologic (long-wave infrared and
visual) and anatomic assessment of skin and subcutaneous tissue
abnormalities and or existing open wounds. The gradiency of the
thermal intensity, not the absolute amount of intensity, is the
important component of the long-wave thermal image analysis that
will allow the clinician to evaluate pathophysiologic events. This
capability is beneficial to the clinician in the prevention, early
intervention and treatment assessments of a developing existing
condition caused by, but not exclusively, wounds, infection,
trauma, ischemic events and autoimmune activity.
[0099] Utilizing temperature values (F..degree., C..degree., and
Kelvin) as the numerical values of LIR thermal heat intensity is
complicated due to the need to have a controlled environment. This
is required since the value of the temperature scales is affected
by ambient temperature, convection of air, and humidity. These
variables would need to be measured and documented continuously if
temperature values were used. Also the emissivity, absorptivity,
reflexivity and transmitability of the skin and subcutaneous tissue
can be affected by skin moisture, scabbing, slough and/or eschar
formation in an open wound.
[0100] To address this problem the imager utilizes the raw data
captured by the microbolometer. This data is utilized in
determining pixel values relating to the intensity of the thermal
energy from the long-wave infrared electromagnetic radiation
spectrum being emitted by the human body. The pixel gradient
intensities are represented for visualization by the grayscale
presentation.
[0101] The pixel values in the grayscale thermal images also vary
with the varying conditions mentioned above and hence the
algorithms proposed in this application use the average pixel value
of the unaffected skin region for that patient on the day the image
was taken as a reference point for all the calculations.
[0102] Combining the above technique with suggested usage of
unaffected skin and subcutaneous tissue in the proximity of an
abnormality of a skin/subcutaneous tissue location as a real time
control helps to minimize the variability and time consuming
requirements in utilizing temperature scales.
[0103] There is a difference in the LIR thermal intensity regions
of the human body. LIR images have a defined pixel intensity range
that is based on the specific usage of an LIR image. In the arena
of skin and subcutaneous tissue LIR thermal gradiency, the range is
within homeostasis requirements to sustain life. The visualization
of pixel intensities is accomplished by the use of a standardized
8-bit grayscale. Black defines cold, gray tones define cool and/or
warm and white defines hot. When the imager is used for capturing
extremely hot or extremely cold regions that fall outside the
thermal range of the imager the pixel values reach the saturation
point and it becomes extremely difficult for the human eye to
differentiate variations in the pixel values.
[0104] This situation can be addressed by utilizing a visualization
technique that increases the pixel values to create a positive
offset to make the image look brighter. In the same manner a
negative offset can be used to decrease the pixel values to make
the image look darker.
[0105] A. Increasing and Decreasing Pixel Value Brightness by
Adding a Positive or Negative Offset to the Raw Pixel Value.
[0106] The positive and negative offset can be utilized to assist
in visualizing the area of the body being imaged. The usage of the
offsets can then be documented as being used at the time the image
is initially taken. The default gray tone that represents the
actual pixel values is the raw data being stored in the data bank
so future analysis can be performed by clinicians at a later time
and/or in another location. The default grayscale data is
accompanied by documentation of the use of either the positive or
negative offset process. This allows for enhanced visualization of
black and white extremes in the grayscale image. The goal is to
visually enhance the image at either the lower or higher side of
the thermal intensity range without altering the original
image.
[0107] Referring to FIG. 2, the thermal imager could be configured
to capture the thermal intensity variation information within a
certain range of thermal intensity. Configuration settings were
carefully chosen such that they capture all thermal intensity
variations between 19.degree. C. (66.2.degree. F.) to 40.5.degree.
C. (104.9.degree. F.), which covers most of the human body's
physiologic thermal intensity range. When the thermal intensity of
an area of interest gets close to 19.degree. C. (66.2.degree. F.),
the pixel values in the grayscale thermal image appear darker and
reach a low saturation point. When the thermal intensity drops
below 19.degree. C. (66.2.degree. F.), the thermal image would
still appear dark but would not get any darker as the low
saturation point has already been reached. Similarly as the thermal
intensity of an area of interest starts increasing, the thermal
image starts looking brighter. As the thermal intensity gets close
to 40.5.degree. C. (104.9.degree. F.), the thermal image reaches
the high saturation point and the pixel values in the grayscale
image reach the maximum value. As the thermal intensity goes beyond
40.5.degree. C. (104.9.degree. F.), even though the thermal
intensity of the area of interest is increasing, the thermal image
would not appear any brighter as the high saturation point has been
reached.
[0108] Even though the thermographic imager can pick up thermal
intensities as low as 19.degree. C. (66.2.degree. F.) the grayscale
thermal image for an area of interest at that thermal intensity
would appear too dark. The human eye is not able to visualize the
variation of the 254 pixel values included in the standardized
grayscale. This might cause problems when thermographic images are
taken on areas of the human body with decreased microcirculation,
(i.e., the fingers, toes, etc.) or areas with cartilage (i.e., the
tip of the nose, ear, etc.). These body locations are usually the
coldest on the skin surface thermal intensity and would appear
darker in the thermal images.
[0109] To solve this problem, a novel technique has been developed
to increase or decrease the brightness of the pixel values by
adding a positive or negative offset to the raw pixel values. The
positive or negative offset allows an enhanced visualization of the
black or white extremes in a grayscale image. The goal here is to
visually enhance the image at either the lower or higher end of the
thermal intensity range without altering the original image.
[0110] With default configuration settings and at a room thermal
intensity of 22.11.degree. C. (71.8.degree. F.), the thermal
intensity range picked up by the thermal imager was as illustrated
in FIG. 2.
[0111] A low saturation grayscale value of 1 was reached at
19.degree. C. (66.2.degree. F.) and the high saturation grayscale
value of 254 was reached at 40.5.degree. C. (104.9.degree. F.),
giving a thermal span of 21.5 degrees. The maximum resolution is
then 0.0846.degree. C. with in the image.
Thermal Span (Thermal intensity range picked up by an
imager)=(Thermal intensity at which the pixels reach the high
saturation value)-(Thermal intensity at which the pixels reach the
low saturation value): Formula
Maximum resolution = ( High saturation temperature - low saturation
temperature ) Resolution of the gray scale image ##EQU00001##
[0112] For an 8-bit grayscale image the resolution is fixed at 254
parts.
[0113] Adding a Positive Offset (Example of Use)
[0114] When a positive offset+20 was added to all the pixels to
make the image look brighter the imager reached the low saturation
grayscale value of 21 at 19.degree. C. (66.2.degree. F.). Since a
value of +20 is added to all the pixels, the grayscale value can
only go as low as 21 instead of I as obtained with default
settings. This lowest grayscale value was obtained at the same
thermal intensity (19.degree. C.) as the low saturation thermal
intensity obtained with default settings. This indicates that
adding an offset will only increase the pixel value making it look
brighter so that small variations in the pixel values could be
visually seen. This does not enable the thermal imager to pick up
thermal intensities lower than what can be read with default
settings.
[0115] With positive offset added, the image appears brighter and
reaches the high saturation value at a thermal intensity lower than
the high saturation thermal intensity obtained with default
settings. The imager reached the high saturation thermal intensity
at 39.degree. C. (102.2.degree. F.) instead of 40.5.degree. C.
(104.9.degree. F.), as obtained with default settings.
[0116] FIG. 3 shows the thermal intensity range that is detected
when a positive offset is added to the default pixel value
configuration setting.
[0117] The thermal span is reduced to 20 degrees instead of 21.5
degrees as obtained with default settings when a positive offset
was added. The maximum resolution increased to 0.0855.degree. C.
which gives more definition to the pixels within the image.
[0118] Adding a Negative Offset (Example of Use)
[0119] Adding a negative offset to the raw signal coming from the
imager makes the thermal image look darker, improving the
visualization of the hot (bright) areas. When an offset of -20 was
added to the original signal the pixel values reached the low
saturation value of 1 at 20.5.degree. C. (68.9.degree. F.) instead
of 19.degree. C. (66.2.degree. F.). Since the thermal images are
saved as unsigned 8-bit grayscale images with pixel values ranging
from 1-254, if the values fall outside this range they would be
mapped to 1 or 254. So when a negative value is added, pixels with
values less than 20 would become negative and were mapped back to 0
so that the pixel values always stay in the range of 1-254.
Similarly on the high end the pixel values reached the highest
saturation value of 234 at 40.5.degree. C. (104.9.degree. F.). With
a negative offset added the highest the pixel values can go up to
is 234 instead of 254. This high saturation occurred at the same
thermal intensity as obtained with default settings.
[0120] FIG. 4 shows the effect of adding a negative offset on the
thermal intensity range that could be picked up by the thermal
imager.
[0121] The thermal span is reduced to 19 degrees giving a maximum
resolution of 0.0855.degree. C. within the image.
[0122] By choosing a suitable offset (positive or negative) value
the visualization of an image is enhanced by increasing the
resolution within the image. This concept has been implemented and
proven by the researched thermal imaging. An offset of 20 was
chosen as an example. This could change based on the requirements.
FIG. 5 below shows a thermal image of a hand taken with default
settings. FIG. 6 below shows an example of the effect on the
thermal image when a positive offset is added to the pixel values
at default settings to improve the visualization of the image.
[0123] B. Defining Pixel Intensity Variations in the Long-Wave
Infrared Image
[0124] To assist the clinician to define the pixel intensity
variations of the long-wave infrared image to see how thermal
intensity is varying across the skin area of images taken, as well
as previous thermal images of the same location, an inventive
technique has been developed that measures the thermal intensity
ratio. This gives the clinician the ability to look at the images
captured with the thermal imager and choose pixel points in the
image utilizing non-zoomed and zoomed presentations of the image
that represent skin and subcutaneous tissue surrounding the area of
interest. The clinician also has the ability to select the tissue
in which an injury/wound exists as shown in FIGS. 7 and 8. The
zoomed capability allows the clinician to be very precise in the
selection of the pixels used to measure thermal intensity. The
zoomed feature is particularly useful because of the complexity of
various wound types. For example, the wound base and periwound can
be disorganized (acute and chronic condition, etc.), organized
(wound resurfacing or repairing, etc.); and/or infected (wound base
infection with and without periwound cellulitis, etc.).
[0125] FIG. 7 shows a non-zoomed thermal image with unaffected and
abnormal selections. The `X` marks represent the unaffected skin,
the asterisk symbol represents the wound base and the circle marks
represent the periwound.
[0126] FIG. 8 shows an original and zoomed thermal image with
abnormal selections. The table in the image shows selected points
on the thermal image with their corresponding grayscale values.
[0127] FIG. 9 shows a schematic representing pixel intensity
recognition (zoomed).
[0128] Pixels with uniform gray color represent the unaffected skin
and subcutaneous tissue. If the pixel value is too high then it can
be an indication of an infection developing in that area. The wound
base is usually colder than the unaffected skin's thermal intensity
and is represented with darker pixels on a thermal image. The pixel
values for a periwound area are usually higher than the wound base
pixel value and less than the pixel value associated with the
unaffected tissue as their thermal intensity falls between the
unaffected skin thermal intensity and the wound base thermal
intensity.
[0129] The display of pixel value associated with each pixel
selection made could help a clinician make a decision on whether an
area of interest is present. This allows the following calculations
to be performed:
[0130] Wound Base to Unaffected Ratio:
Wound base to unaffected ration = ( Average of all the pixel values
from the wound base region ) ( Average of all the pixel values from
the unaffected region ) ##EQU00002##
[0131] Wound base regions are usually colder than the unaffected
skin thermal intensity, causing the pixel values for the wound base
regions to be lesser than the pixel values for the unaffected skin
regions in an LIR image.
[0132] If the wound base to unaffected ratio is less than 1, it is
an indication that the wound base is colder than the unaffected
regional tissue. If the ratio is greater than 1, it is an
indication that the wound base area is hotter than the regions
selected as unaffected skin area. In summary, the closer the value
gets to 1, the closer the wound base area is getting to unaffected
skin.
[0133] Periwound to Unaffected Ratio:
Periwound to unaffected ratio = ( Average of all the pixel values
from the periwound region ) ( Average of all the pixel values from
the unaffected region ) . ##EQU00003##
[0134] If the periwound to unaffected ratio is less than 1, it
indicates that the periwound is colder than the unaffected skin
area. If the ratio is greater than 1, it is an indication that the
periwound area is hotter than the regions selected as unaffected
skin area. In summary, the closer the value gets to 1, the closer
the periwound area is getting to unaffected skin.
[0135] Periwound to Wound Base Ratio:
Periwound to wound base ratio = ( Average of all the pixel values
from the periwound region ) ( Average of all the pixel values from
the wound base region ) ##EQU00004##
[0136] The ratio greater than 1 indicates that the periwound region
is hotter than the wound base region and the ratio less than 1
indicates that the wound base region is hotter than the periwound
region. In summary, the closer the ratio gets to 1, the closer the
wound base and periwound values get to each other.
[0137] By monitoring these ratios the clinician could get a better
idea on the status of the wound.
[0138] C. Maintaining Separation of the Imager from Target
[0139] Long-wave infrared and visual images must be consistently
taken at a predetermined distance, typically 18 inches. This
capability allows measurements to be obtained by
length.times.width, by linear measurement, and by encirclement of
the area of interest and or wound. This information is considered
to be the gold standard of the wound care industry in determining
the progression or regression of abnormalities.
[0140] Thermal and visual cameras are used for capturing images of
areas of interest, such as wounds in a real time fashion (i.e.,
bedside or outpatient clinic). Cameras are built so that they can
communicate with computer via a USB connection and capture both
visual and thermal images by clicking the trigger button on the
camera.
[0141] All the images need to be captured at a certain distance
from the body part and a standard distance of 18 inches between the
camera and the body part was found in testing done to date to be an
ideal distance. Several methods were used in order to measure this
distance.
[0142] As a first attempt an antenna of length 18 inches was placed
on the camera core that could be extended out. When the end of the
antenna touched the body part the standard distance was known to
have been attained, indicating that the camera is ready for
capturing images. The adverse effects of using an antenna for
measuring the distance were that the antenna would be touching the
body part giving rise to possible risk of contamination, and also
that the antenna comes into the field of view when the image is
being captured causing problems with visualization.
[0143] To overcome these problems the antenna method was replaced
with a more sophisticated method using ultrasonic sound waves. An
ultrasonic transducer placed on the camera core would release
ultrasonic sound waves for transmission in the desired path and
when these waves hit the target, which would be the body part in
our case, and ultra sonic sound waves would be reflected back from
the target in the transmission path. The received ultrasonic sound
waves can then be converted into an electrical signal that can be
processed by a processor to provide distance information. The
distance can be computed by using the time period from the middle
time value of the received electrical signal to the middle time
value of the transmitted signal. Whenever this distance equals the
standard distance of 18 inches a reduced audible noise will be
generated, indicating that the camera is ready to capture an
image.
[0144] Even though the ultrasonic sound wave method has been proven
to be successful and has been used in various applications to date
for measuring the distances, it was never used in the medical field
at bedside as a tool for capturing visual and thermal images.
[0145] Limitations of using the ultrasonic method included the
complexity of wiring and the size of the apparatus used for
measuring the distance and then displaying it so that the end user
can see how far the camera is from the target. The other major
limitation arose with the presence of an object in between the
camera and the target. When there is an object in the path, part or
all of the waves will be reflected back to the transmitter as an
echo and can be detected through the receiver path. It is difficult
to make sure that the received ultrasonic sound waves were actually
reflected by the target and not by any other object in the
path.
[0146] The ultrasonic measuring of the distance was replaced with
the use of two Class 1 Laser LED lights. Two Class 1 A, or of less
strength, lasers and/or LED modified lights are used in this
method. These lasers emit narrow light beams as opposed to diffused
light. They are placed on either side of the camera lens. When the
distance between the camera and the target is less than 18 inches
the lights coming from these lasers fall on the target as two spots
separated by a distance and this distance will keep decreasing as
the camera is moved toward from the target. When the distance
between the camera and the target equal 18 inches the lights from
these two light sources will coincide, indicating that the focus
point has been achieved and that the camera is ready for capturing
images. The distance between the two light beams starts increasing
again when the distance between the camera and the target increases
to the standard 18 inches.
[0147] FIG. 10 explains the above embodiment in more detail, where
IFR represents the long wave infrared microbolometer and D
represents the visual digital camera, and L represents the laser
lights.
[0148] Depending on how far the laser lights are going to be from
the microbolometer and the distance between the microbolometer and
the target the angles at which the lasers need to be inclined will
change.
[0149] The digital camera `D` is also going to be placed at around
1.5 inches away from the long-wave infrared microbolometer and in
order to make both the digital and the long-wave infrared
microbolometer to have the same focus point and field of view the
digital camera needs to be inclined at an angle.
[0150] The experimental setup of FIG. 11 that was used in order to
determine the angle of inclination is as shown.
[0151] FIG. 12 is a representation of an embodiment that uses 18
inches as the desired distance in a clinical setting. By changing
the angles of the Class 1 Lasers this distance can be increased or
decreased to meet other needs or requirements determined by the
clinician.
[0152] D. A Consistent Technique to Obtain Wound Measurement Length
and Width Linearly Using a Thermal Image
[0153] To assist clinicians with maintaining accuracy and
consistency when measuring a wound, a novel technique has also been
developed to obtain consistent linear wound measurements (length
and width) using a thermal image. It allows a clinician to follow a
standard of care to determine the progression and regression of the
wound by measuring length and width and area.
[0154] To be able to obtain the measurements of a wound from an
image the number of pixels available per centimeter or per inch in
that image needs to be known. When images are always taken from a
standard distance the number of pixels per inch in that image
always remain constant, and they change with the change in the
separation distance between the imager and the target.
[0155] The imager has been designed such that the separation
distance between the imager and the target is always maintained at
18 inches. Several techniques like using a measuring tape, using
ultrasound and using Class 1 lasers have been tried and tested to
date to maintain this standard distance. The final version of the
imager makes use of two Class lasers mounted inside the imager at
an angle such that the laser beams emitted from these two lasers
always converge at 18 inches from the front of the camera.
[0156] For an image taken at a distance between the object being
imaged and the imager that is exactly 18 inches there would be in
the image approximately 40 pixels per inch. This distance can be
changed, but at each distance the number of pixels needed to equal
1 cm or 1 inch must be measured and tested. The selected distance
must be noted to maintain reproducibility. For the calculation of
length and width of the wound, when a line is drawn across the area
of interest by measuring the number of pixels covered across this
line and using a conversion formula the measurement in pixels could
be converted into inches or centimeters. For an image taken at 18
inches from the target, a line that is 40 pixels in length would be
approximately 1 inch on the measuring scale and using the inch to
centimeter conversion the length could then be converted into
centimeters.
[0157] Algorithm for Measuring Length and Width of an Area of
Interest (in Centimeters)
[0158] Draw a line across the image that represents the length or
width of the area of interest that needs to be measured.
[0159] Note the x and y coordinates of the starting and ending
points of this line.
[0160] If (x1,y1) represent the x and y coordinates of the starting
point of the line and (x2,y2) represent the x and y coordinates of
the end point of the line then the distance between these two
points (length of the line in pixels) can be measured as:
Length ( or width ) in pixels = ( x 2 - x 1 ) 2 + ( y 2 - y 1 ) 2
##EQU00005## Length ( or width in inches = Length in pixels 40
Length ( or width ) in centimeters = Length in pixels 15.7480
##EQU00005.2##
[0161] As per Minimum Data Set (MDS) Version 3.0, it is recommended
that the length of a wound is always measured as the longest length
drawn from head to toe and width is measured as the widest width
drawn side to side perpendicular to the length: The x or y
coordinates of the end point of the line representing the length or
the width line could be adjusted to make sure the lines are exactly
vertical or horizontal which would in turn make them perpendicular
to each other.
[0162] Using the length and the width measurements
(length.times.width) area could be calculated.
[0163] By monitoring the thermal images taken on day to day basis,
and by measuring the length and the width for the area of interest
each day, the status of the wound could be monitored to see whether
there has been a progression or regression in the status. FIG. 13
shows the length and width measurement in centimeters obtained for
an image with an area of interest on a heel.
[0164] E. Highlighting the Wound Base, Periwound and Unaffected
Regions to Measure and Calculate the Square Areas Thereof by Using
the Number of Pixels Highlighted
[0165] A novel technique has been developed that gives the
clinician the ability to highlight a wound base, periwound or
unaffected regions and to measure the area in square centimeters.
This will assist the clinician in looking at the overall status of
the wound, and evaluating its progression or regression.
[0166] The total number of pixels enclosed within the highlighted
area could be used for calculating the area of the region
selected.
[0167] A test target of size 1.5 inch.times.1.5 inch was used. With
the imager at 18 inches from the test target, images were
captured.
[0168] The area of test target=1.5 inch.times.1.5 inch=2.25 square
inches or 3.81 cm.times.3.81 cm=14.5161 square cm.
[0169] For an image taken at 18 inches from the target there would
be approximately 40 pixels per inch. So there would be
approximately 60 pixels in 1.5 inches.
[0170] The area of the test target obtained from the image=60
pixels.times.60 pixels=3600 pixels. A total of 3600 pixels were
enclosed inside the area of the test target. So 3600 pixels14.5161
square cm
[0171] For an unknown area of interest, if "Y" is the number of
pixels enclosed inside that area then the surface area in square
centimeters for that region would be equal to:
Area in square centimeters for the highlighted region = ( Y .times.
14.5161 ) 3600 ##EQU00006##
[0172] For the region highlighted as the wound base, area in square
cm's and average of all the pixel values falling inside the
highlighted region are calculated and displayed in the picture as
shown in FIG. 14.
[0173] Periwound area represents the area surrounding the wound
base. By highlighting the area that includes the wound base and the
periwound area surrounding it as shown in FIG. 15, and by counting
the number of pixels enclosed in that region, the area of the
highlighted region could be calculated in square centimeters. The
periwound area could then be obtained by subtracting the wound base
area from the area that includes both the periwound and the wound
base areas.
[0174] By including the unaffected skin and subcutaneous tissue
surrounding the wound in the highlighted area of interest, the
unaffected area could be calculated in square centimeters. The
unaffected area could then be obtained by subtracting the wound
base and periwound area from the region selected that includes
unaffected, periwound and the wound base areas.
[0175] FIG. 16 below shows the calculations displaying the
highlighted unaffected area and the various calculations obtained
from the highlighted regions.
[0176] F. Obtaining the Average Pixel Value and the Plus/Minus
Variance by Encircling the Area of Interest/Wound
[0177] By utilizing the novel techniques above, not only can the
area be calculated, but simultaneously the average pixel value of
each area can be calculated. This will allow the clinician to
evaluate the status of the area of interest or wound not only in
micro (focused technique, above) but also in the macro using the
technique described below. The combination of these two assessments
will give a better overall understanding of the areas of interest
where the abnormality or wound has been identified. From this
average data the ratio concept discussed above can also be used to
evaluate the macro (overall) look at an area of interest or wound,
specifically if the wound is becoming organized, i.e., is it
improving, becoming infected, or regressing (getting worse). See
Table 1 below.
TABLE-US-00001 TABLE 1 Summarizing the results obtained from the
highlighted nomzal, periwound, and wound base regions Normal
Periwound Wound Base Area in sq. cm 29.03 16.66 7.71 Average pixel
value 125.17 103.82 61.09 Minimum and [Various range] [Various
range] [Various range] maximum pixel values
[0178] Some of the other measurements that could be done to keep
track of the status of an area of interest include calculating the
average, minimum and maximum of all the pixel values falling inside
the highlighted area.
Average pixel value = Sum of pixel values for all the pixels that
fall inside the highlighted area of interest Total number of pixels
falling inside the highlighted area ##EQU00007##
[0179] For a highlighted area of interest a histogram can be
generated to provide graphical representation of distribution of
pixel values within that area.
[0180] Algorithm for generating histograms.
[0181] Highlight the area of interest for which a histogram needs
to be generated
[0182] Determine the total number of bins/buckets into which the
data needs to be divided into. There is no best number of bins, and
different bin sizes can reveal different features of the data.
[0183] Bin size can be calculated as
Bin size = Maximum value ~ Minimum value Total number of bins
##EQU00008##
[0184] For a thermal image the pixel values always range between 0
and 255.
[0185] Create an empty array of size equal to the total number of
bins.
[0186] Check to see if a pixel falls inside the highlighted area of
interest and if it does note the pixel value.
[0187] The bin number into which this pixel value falls under can
be calculated using the formula.
Bin number = Pixel value - Minimum value Bin Size ##EQU00009##
[0188] Increment the value of the array at the index [Bin
number-1], since arrays are zero based, by one.
[0189] Repeat the steps 5-7 for all the pixels in an image.
[0190] After checking all the pixels in an image, plot the array to
generate a histogram.
[0191] Clinical Significance of Histograms.
[0192] Distribution of pixel values as projected by the histograms
for a highlighted area of interest provides more in depth
information about the signature of a wound. If the histogram plot
is more spread out it indicates there is a large variation in the
pixel values and hence temperatures within the highlighted area as
shown in FIG. 17. As the plot starts getting more and more narrow
it is an indication that all the pixels inside the highlighted
portion are getting close to each other and the temperature inside
the highlighted portion is starting to get saturated towards a
single temperature value. If the saturation occurs at a higher
pixel value then it is an indication that all the pixels inside the
highlighted portion are getting very hot compared to the selected
normal reference point. Similarly if the saturation occurs at a
very low pixel value then all the pixels inside the highlighted
area are getting very cold. FIG. 17 shows some sample histograms
generated for an image with a highlighted area of interest.
[0193] G. Creating Profile Lines in and Through an Area of
Interest/Wound and Comparing with Profile Lines Trough Reference
Areas
[0194] A novel feature has been developed to assist a trained
clinician to better track a wound by utilizing the ability to plot
profile lines through the wound. These plots show the variation in
the pixel values across the wound. Since the thermal intensity is
directly related to the grayscale pixel values in an image, these
plots can be used to monitor how the thermal intensity is varying
across an area of interest or wound. This allows the clinician to
dissect the wound in precise fashion so the pathophysiologic status
of the wound can be assessed and quantified.
[0195] Profile lines can be plotted by simply drawing a line across
the area of interest. FIG. 18 below shows an example of the profile
line generated by drawing a line across the wound present on the
heel. As seen in the plot there is a huge drop in the pixel
value/thermal intensity across the wound base region and the value
starts increasing as the line is moving away from the wound base
and entering areas with unaffected skin tissue.
[0196] As the wound starts healing the difference between the pixel
value for the unaffected tissue and the pixel value from the wound
base starts decreasing and hence the drop seen in the graph starts
decreasing indicating that the wound is healing and is starting to
get close to the unaffected skin tissue.
[0197] If the drop in the pixel values starts increasing, when
plots are generated for images taken on timely basis then it is an
indication that the wound is deteriorating and that the clinician
needs to turn to strategies to facilitate wound healing.
[0198] Algorithm for Generating the Profile Lines
[0199] Draw a line across the area of interest for which the
profile lines need to be plotted.
[0200] Record the X and Y locations of the starting and end points
of the profile line. Let (x1, y1) represent the coordinates of the
starting point and (x2, y2) represent the coordinates of the end
point.
deltaX=absolute value of (x2-x1); deltaY=absolute value of
(y2-y1)
length of the line=L= {square root over
((x2-x1).sup.2+(y2-y1).sup.2)}
x_increment=deltaX/L
y_increment=deltaY/L
[0201] Round off L to the nearest integer and then increment by 1;
L=L+1
[0202] Create a new array to hold the pixel values that fall across
the profile line. Let us call this array.
[0203] `Pixel Values`.
[0204] Pixel_values(1)=pixel value of the image at the location x1,
y1. Add the x_increment and y_increment to the original x1 and y1
respectively and use these as new values for x1 and y1. So x1=round
(xi+x_increment) y1=round (y1+y_increment)
[0205] Create a new counter variable, let us call it `i`
[0206] i=1;
[0207] While ((i<L) && (x1, y1 fall within the size of
the image)
Pixel_values (i+1)=pixel value of the image at the location
x1,y1;
x1=round (x1+x_increment);
y1=round (y1+y_increment);
i=i+1;
[0208] End
[0209] The array `Pixel_values` should contain values of all the
pixels that represent the profile line.
[0210] Plotting the values in the array `Pixel_values` gives the
plot for the profile line drawn across the area of interest (as
shown in the figure above).
[0211] Images taken using a thermal imaging camera can be analyzed
and tracked to monitor the status of wounds.
[0212] Profile lines provide a tool for monitoring variations in
pixel values and hence the temperatures across the abnormal areas
of interest. These variations can be compared against the pixel
value representing unaffected region for that patient by selecting
a region on the image that represents unaffected skin.
[0213] Comparing the Profile Line with the Reference Line
Representing the Unaffected Skin for that Patient.
[0214] Comparing the pixel values of the pixels falling across the
profile line with the reference pixel value that represents
unaffected skin for that patient gives a measure of how close or
far away the profile line pixel values are from the selected
reference line.
[0215] For selecting unaffected regions a circle can be drawn on
the image that comprises of only the unaffected pixels and does not
include any abnormalities or the background. Once a circle has been
drawn representing unaffected skin for the patient, average of all
the pixels falling inside the circle can be calculated as
follows:
Average Normal pixel value = Sum of all the pixels that fall inside
the circle representing Normal Total number of pixels inside the
circle ##EQU00010##
[0216] To determine whether a pixel falls inside a circle of radius
`r` calculate the distance between the center of the circle and the
coordinates of the pixel point using the formula
Distance= {square root over ((x2-x1).sup.2+(y2-y1).sup.2)}
where (x1, y1) represent the X and Y coordinates of the center of
the circle and (x2,y2) represent the X and Y coordinates of the
pixel.
[0217] If the distance is less than the radius of the circle then
that pixel falls inside the circle representing unaffected skin
area.
[0218] Once the average normal pixel value has been calculated this
value can be plotted on the chart along with the profile line as
shown in the FIGS. 20 and 21.
[0219] By comparing the profile line with the normal line the
status of the area of interest can be tracked. As the profile line
gets closer to the reference line it indicates that the area of
interest is improving and is getting closer to the normal skin
characteristics.
[0220] The portions below the reference line represent the segments
of the profile line where the pixel values are lower (colder) than
the selected normal reference point. Similarly the points falling
above the reference line represent the portion of the profile that
is hotter than the selected normal reference.
[0221] Once a normal reference point has been chosen and a profile
line has been drawn several parameters can be calculated to compare
the profile line signature with the reference line signature. By
tracking how these values change on day to day basis the status of
the wound could be tracked.
[0222] Some of the factors that could be calculated to compare the
profile line with the reference line include area above and below
the reference line, maximum rise and drop, average rise and drop
from the reference line etc.
[0223] The area calculations also give a measure of the portion of
the profile line that falls above or below the normal reference
line. The area that falls above the reference line indicates the
regions that have a pixel value higher that the reference point and
hence are at a higher temperature. The area below the reference
line shows the portion of the profile line that has temperatures
lower than the selected reference.
[0224] The areas can be calculated using the Trapezoidal rule of
calculating area under the curve.
[0225] Calculating Area Above and Below the Reference Line
[0226] The area between the graph of y=f(x) and the x-axis is given
by the definite integral in FIG. 22 (Reference:
http://www.mathwords.com/alarea_under_a_curve.htm) This formula
gives a positive result for a graph above the x-axis, and a
negative result for a graph below the x-axis.
[0227] Note: If the graph of y=f(x) is partly above and partly
below the x-axis, the formula given below generates the net area.
That is, the area above the axis minus the area below the axis.
[0228] The trapezoidal rule (also known as the trapezoid rule or
trapezium rule) is an approximate technique for calculating the
definite integral as follows
.intg. a b f ( x ) x .apprxeq. .DELTA. x 2 * ( f ( x 0 ) + f ( xn )
+ f ( x 2 ) + + f ( x ( n - 1 ) ) ) ) ##EQU00011##
[0229] Where
.DELTA. x = ( b - a ) n , x 0 = a , x 1 = a + .DELTA. x , x 2 = a +
2 .DELTA. x xn = a + n .DELTA. x = b ##EQU00012##
and `n` is the number of equal length subintervals into which the
region [a, b] is divided into.
[0230] To calculate area relative to the Normal line, instead of
x-axis, pixel values relative to the selected normal need to be
calculated which equal to the actual pixel value-the selected
normal value.
[0231] If relative pixel value is positive it indicates that the
point falls above the normal line and if negative it falls below
the normal line. Whenever the relative pixel value across the curve
goes from positive to negative or vice versa it is an indication
that there has been a crossover across the normal line. The
algorithm for computing the area above and below the normal line
can be summarized as follows:
[0232] Calculate relative pixel values
[0233] Find out where the crossover points occur
[0234] Split the curve into positive and negative regions
[0235] Calculate area for each region separately using the
Trapezoidal rule
Finally, combine all positive areas to obtain area above normal
line and all the negative areas to obtain the area below the normal
line
[0236] FIG. 23 shows a plot of a sample profile line and a normal
line. As shown in the figure the sample profile lines crosses the
normal line at three points dividing the curve into three regions.
Regions 1 and 3 fall above the normal line and have positive
relative pixel values whereas the region 2 falls below the normal
line and has negative relative pixel values.
[0237] To calculate the area above and below the normal for the
sample plot the area for the three regions need to be calculated
individually using the Trapezoidal rule.
Area for the region 1 = .intg. a b 1 f 1 ( x ) .apprxeq. .DELTA. x
2 * ( f 1 ( a ) + f 1 ( b 1 ) + 2 * ( f 1 ( x 1 ) + f 1 ( x 2 ) + +
f 1 ( x ( n - 1 ) ) ) ) ##EQU00013##
[0238] Where f1(x) defines the curve in region 1
.DELTA. x = ( b 1 - a ) n , x 1 = a + .DELTA. x , x 2 = a + 2
.DELTA. x xn = a + n .DELTA. x = b 1 and ` n ` is the
##EQU00014##
number of equal length subintervals into which the region [a,b1] is
divided into.
Area for region 2 = .intg. b 1 b 2 f 2 ( x ) .apprxeq. .DELTA. x 2
* ( f 2 ( b 1 ) + f 2 ( b 2 ) + 2 * ( f 2 ( x 1 ) + f 2 ( x 2 ) + +
f 2 ( x ( n - 1 ) ) ) ) ##EQU00015##
[0239] Where f2(x) defines the curve in region 2
.DELTA. x = ( b 2 - b 1 ) n , x 1 = b 1 + .DELTA. x , x 2 = b 1 + 2
.DELTA. xn = b 1 + n .DELTA. x = b 2 ##EQU00016##
and `n` is the number of equal length subintervals into which the
region [b1, b2] is divided into.
[0240] The area for this region would be negative indicating that
it falls below the normal line.
[0241] Area above the Normal line can be obtained by adding areas
under regions 1 and 3=
.intg..sub.a.sup.b1f1(x)+.intg..sub.b2.sup.bf3(x)
Area below the Normal line=Area under the region
2=.intg..sub.b1.sup.b2f2(x)
[0242] By counting exactly how many number of pixels fall above or
below the reference line the percentage of profile line that falls
above or below the profile line can be calculated as follows:
Percentage of profile line that falls above the reference line = (
Number of pixels that fall above the reference line ) * 100 Total
number of pixels across the profile line Percentage of profile line
that falls below the reference line = ( Number of pixels that fall
below the reference line ) * 100 Total number of pixels across the
profile line Percentage of profile line that falls along the
reference line = ( Number of pixels that fall on the reference line
) * 100 Total number of pixels across the profile line
##EQU00017##
[0243] Maximum rise above the reference line gives the maximum
positive difference in the pixel values between the profile line
and the reference line. A rise in this value indicates that the
temperature for some of the pixels along the profile line is
getting much hotter than the reference value and decrease in this
value indicates that the maximum difference between the profile
line pixel values and the reference line pixel values is decreasing
and that the profile line is getting closer to the reference
line.
[0244] Similarly Maximum drop below the reference line can be
calculated as the maximum negative difference in the pixel values
between the profile line and the reference line. An increase in the
maximum drop indicates that the pixels on the profile line are
colder than the average reference pixel value.
[0245] Average rise and average drop can also be used as factors
for comparing the profile lines with the reference line. Formulae
for calculating average rise and average drop are as follows:
Average rise above the reference line = Sum of all the pixels that
fall above the reference line Total number of pixels that fall
above the reference line Average fall below the reference line =
Sum of all the pixels that fall below the reference line Total
number of pixels that fall below the reference line
##EQU00018##
[0246] Slopes: Calculating slopes for the profile lines gives
information about how often the temperature varies along the
profile line. A slope line can be drawn on the profile line every
time there has been a significant change in the pixel value
(temperature). A positive slope indicates an increase in
temperature and a negative slope indicates a drop in the
temperature. The steepness of the slope lines indicates the amount
of variation in temperatures. The steeper the lines the larger the
variation is temperatures and the more irregular the profile line
is.
[0247] An algorithm for calculating slopes and generating slope
lines across the profile line can be summarized as follows:
[0248] Select a suitable value for slope variance, a value which
indicates how much of a difference in pixel values between two
points on the profile line is considered as a signification
change.
[0249] Consider the starting point of the profile line as the
starting point of the first slope line. Starting from this point
and by moving along the profile line calculate the difference
between the current pixel value and the pixel value at the starting
location. If the difference is greater than or equal to the slope
variance, the point at which the difference exceeds the slope
variance becomes the end point for the slope line.
[0250] Draw a line on the profile line joining these two
points.
[0251] Slope for this line can be calculated as follows:
[0252] If (x1, y1) represents the x and y coordinates of the
starting point and (x2, y2) represent the coordinates of the end
point of the slope line then the slope for this line can be
calculated as
Slope = ( y 2 - y 1 ) ( x 2 - x 1 ) ##EQU00019##
[0253] Save this slope value in an array.
[0254] Make the end point of the first slope line as the start
point for the next slope line to be generated and repeat step 2 to
determine the new end point.
[0255] Once the start and end points of the slope lines is
established plot the slope line on to the profile line and then
calculate and save the slope values.
[0256] Repeat the process until the end of profile line is
reached
[0257] FIG. 21 shows a slope line plotted on to the profile line
with a slope variance of 12.
[0258] These are some of the factors that can be calculated from
the profile line and reference line plots that help define the
signature of the area of interest.
[0259] All the activity done by the clinician on the images can be
recorded and saved in a database. The information can be retrieved
on a later date to see which regions were selected as area of
interest on that particular day, and to see what changes have
occurred and how the results have changed with time. This novel
approach will enable a trained clinician to better evaluate the
area of interest/wound of the skin and subcutaneous tissue in a
standardized and reproducible format.
[0260] The benefits related to using this advancement in long-wave
infrared thermal imaging spans improvement in potential care,
fulfilling regulatory requirements and fiduciary responsibility by
reproducible and standardized documentation and cost savings
secondary to the ability of clinicians to formulate appropriate
individualized care plans for prevention, early intervention and
treatment of abnormalities of the skin and subcutaneous tissue.
[0261] H. Using the Profile Line Plot to Interpret Wounds
[0262] Once a profile line is drawn on the image across the area of
interest a profile line plot can be generated using the algorithm
outlined above. The plot can then be used to determine where on the
profile line a drop or rise in the pixel value (temperature)
occurs. The profile line plot can be made interactive so that when
the user clicks on the plot the corresponding location on the image
can be highlighted and hence making it easier to interpret. The
algorithm for implementing this can be briefly summarized as
follows:
[0263] 1. Generate an interactive plot for profile line using tools
like Telerik.
[0264] 2. Create a chart item click event for the plot so that when
the user clicks on the profile line plot the x and y values of the
click point are recorded.
[0265] 3. The X axis value at the click point (saved as `index`)
shows how far away the point falls from the start point of the
profile line. The Y value gives the actual pixel value at the
point.
[0266] 4. To locate this point on the profile line drawn on the
image, the actual X and Y coordinates on the image need to be
determined. The X and Y coordinates of the click point can be
obtained as follows:
[0267] 5. Calculate the length of the profile line using the start
and end coordinates of the profile line.
[0268] 6. If (XI,YI) represents the coordinates of the starting
point of the profile line on the image and (X2,Y2) represent the
end point then the length can be calculated as
[0269] 7. length of the line=L= {square root over
((x2-x1).sup.2+(y2-y1).sup.2)}
[0270] 8. deltaX=absolute value of (X2-X1); deltaY=absolute value
of (Y2-Y1)
[0271] 9. x_increment=deltaX/L; y_increment=deltaY/L
[0272] 10. if (x_increment>0 && y_increment<0)
[0273] {
[0274] index=L-index:
[0275] }
[0276] 11. The X and Y coordinates of the point that represents the
click point can then be obtained as X=X1+(index*x.sub.--
increment); Y=Y1+(index*y_increment);
[0277] 12. Draw a string on the image at the X and Y coordinates
from the previous step to indicate the click point
[0278] Similar technique can be used to determine where a point on
the image falls on the profile line. The algorithm for doing this
can be outlined as follows:
[0279] 1. Add a Mouse down click event for the image.
[0280] 2. Note the X and Y coordinates of the point where the user
clicked on the image.
[0281] 3. Check whether this point falls on the profile line
[0282] 4. If the point falls on the profile line calculate the
distance between the start point of the profile line and the point
where the user clicked.
[0283] 5. This distance indicates how far the point falls on the
plot from the start point of the graph.
[0284] 6. Draw on the graph to indicate this point.
[0285] FIG. 24 shows a profile line drawn on the image of a hand
and FIG. 25 shows the profile line plot. The X mark on the graph
and the image indicates the user's selected point.
[0286] While this invention has been described with respect to at
least one embodiment, the present invention can be further modified
within the spirit and scope of this disclosure. This application is
therefore intended to cover any variations, uses, or adaptations of
the invention using its general principles. Further, this
application is intended to cover such departures from the present
disclosure as come within known or customary practice in the art to
which this invention pertains and which fall within the limits of
the appended claims.
* * * * *
References