Introduction For real surfaces, R(T) = ?(T)

Introduction

 

Thermal imaging is a fast, passive
and non-invasive medical imaging modality used to measure and analyse
physiological functions and pathology related to the thermal homeostasis and
temperature of the body. The
technique involves the detection of infrared radiation that can be correlated directly
with the temperature distribution of a defined body region. The
human body is capable of maintaining a constant temperature that can be
different from the ambient temperature. The body temperature is preserved
within fine limits typically at 37°C ± 1°C, and it is essential for the normal functioning
of the human body. Change in the body temperature by a few degrees is
considered to be a clear indication of probable disease. Thermal imaging has
been used successfully in the diagnosis of breast cancer, diabetes neuropathy
and peripheral vascular disorders. It has also been used to detect problems related
to gynaecology, kidney transplantation, dermatology, heart, neonatal
physiology, fever screening and brain imaging. With the advent of modern
infrared cameras, data acquisition and processing techniques, it is now
possible to have real time high resolution thermographic images.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

 

 

Thermal
Radiation

 

All objects with temperature above
absolute zero emit electromagnetic radiation, which is known as infrared
radiation or thermal radiation. Wavelength of this radiation lies within a
range of 0.75–1000 ?m. This wide range can be further subdivided into three
smaller groups as near infrared – NIR (0.76–1.5?m), medium infrared – MIR
(1.5–5.6?m) and far infrared – FIR (5.6–1000?m). According to thermal radiation
theory, black body is considered as a hypothetical object that absorbs all
incident radiation and radiates a continuous spectrum. The Stefan–Boltzmann law for total
radiation emitted from a perfectly black body is given by

R(T) = sAT4

where

R(T) is the total power radiated into a
hemisphere

? is
the Stefan–Boltzmann constant (5.67 × 10-8 W m-2 K-4)

A is the effective radiating area of the body

T is the absolute temperature of the radiating surface

 

For real surfaces,

R(T) = ?(T) sAT4

where ?(T) is the
emissivity of the emitting surface at a ?xed wavelength and absolute
temperature T

For a thermal black body emissivity is unity, but for
real surfaces emissivity is always less than unity.

 

The rate of heat loss between two surfaces at T1 and
T2 is

W = A s ?(T1) T1 4
– ? (T2) T2 4

This radiant heat loss forms the basis of thermal imaging.

 

 

Emissivity of different
human tissues at 40 0C in infrared wavelength

Tissue

Emissivity

Black skin (3–12?m)

0.98 ± 0.01

White skin (3–14?m)

0.97 ± 0.02

Burnt skin (3–14?m)

0.97 ± 0.02

Epicardium (fresh:
0.5 h) 3?m

0.85

Epicardium (fresh:
0.5 h) 5?m

0.86

Epicardium (9 days at
-20 0C)

0.99

Pericardium (3?m)

0.88

Pericardium (5?m)

0.94

Pericardium (9?m)

0.95

 

Infrared emissions from human skin
occur between 2 and 20?m with maximum emission around 10?m. Emissivity of human
skin is almost constant and its value is 0.98 ± 0.01 for wavelength range of
2–14?m. Steketee has found that the emissivity of white skin, black skin and
burnt skin is the same and it is independent of wavelength.

 

Instrumentation

An infrared scanning device is used to convert
infrared radiation emitted from the skin surface into electrical impulses that
are visualised in colour on a monitor. This visual image graphically maps the
body temperature and is referred to as a thermogram. The spectrum of colours
indicates an increase or decrease in the amount of infrared radiation that is
emitted from the body surface. Since there is a high degree of thermal symmetry
in the normal body, subtle abnormal temperature asymmetry can be easily
identified.

A typical imaging device consists of a system for
collecting radiation from a well-defined field of view and a detector that
transduces the radiation focused onto it into an electrical signal. From 1960
to about 1975, images were made with scanning systems that used single IR
detectors. The period from 1975 to 1995 saw the commercial development of the SPRITE
(signal processing in
the element) detector,
linear arrays and 2D arrays. Since 1993 focal-plane array (FPA) detectors have become
available.

 

Single photon
detectors

These
are devices of semiconductor-type in which the photon absorption results in the
freeing of bound electrons or charge carriers in proportion to the intensity of
the incident radiation. In order for the energy gaps to be small enough to
allow the detection of radiation beyond 10?m, mixed crystal detectors such as
CdHgTe (CMT) and PbTe have been used.

The
parameters by which photon IR detectors are usually specified are responsivity,
noise, detectivity, cut-off wavelength and time constant. Responsivity (R)
is the ratio of the output voltage to the radiant input power, expressed in
volts per watt. Since the output voltage due to incident infrared radiation is
a very small fraction approximately 10-5 of the DC bias voltage
across the detector, the responsivity is measured by exposing the detector to
chopped radiation from a calibrated source at 500 K and measuring the
alternating voltage component at the chopping frequency. Responsivities (R,
500 K) are typically 104?105 V W-1 for CMT
detectors operated at ?196°C. Noise, together with responsivity determines the
ability of a detector to detect small input signals. This is specified in volts
per hertz at one or a number of frequencies or as a noise spectrum.

Detectivity
(D) is given by:

D = 1/NEP

where NEP is the noise equivalent
power, which is the rms value of the modulated sinusoidal radiant power that falls
upon a detector that will give rise to an rms signal voltage (Vs)
equal to the rms noise voltage (Vn) from the detector. For
many photon detectors, the NEP is directly proportional to the square root of
the area of the detector and it becomes appropriate to use a normalised
detectivity D* given by

D* = DAd1/2 (?f) = (Vs/ Vn)
Ad(?f) 1/2/ W

where

Ad is the area of detector

?f is the frequency
bandwidth of the measuring system

W is the radiation power incident on
the detector (rms)

 

D* varies with the wavelength of the radiation and the
frequency at which the noise is measured. As a figure of merit, D*
enables a theoretical maximum detectivity to be calculated that would apply
when performance is limited only by noise due to the fluctuation of the
background radiation. The detector time constant ? is the time between incident
radiation being cut off and the output of the detector dropping by 63%. Typical
time constants range from a fraction of a microsecond for CMT detectors to a
few microseconds for InSb detectors.

 

Thermographic
Scanning Systems

The detector forms a part of an
imaging system whose performance is generally specified in terms of temperature
resolution, angular resolution and field of view. Temperature resolution is a
measure of the smallest temperature difference in the scene that the imager can
resolve. It depends on the efficiency of the optical system, the responsivity
and noise of the detector and the SNR of the signal-processing circuitry.
Temperature resolution can be expressed in two ways: noise equivalent
temperature difference (NETD), which is the temperature difference for which
the SNR at the input to the display is unity and minimum resolvable temperature
difference (MRTD), which is the smallest temperature difference that is
discernible on the display. Most of the medical thermography has been carried
out with systems that have MRTD between 0.1 and 0.3 K. Angular resolution is
typically 1–3 mrad but can be as small as 0.5 mrad.

In an imaging system that is capable
of transmitting and focusing infrared radiation, the scene is viewed by an
optical lens. A high value of IR refractive index is advantageous in the lens
design but materials that have high refractive indices tend to have low
transmittances. This high reflective loss may be eliminated by antireflection
coatings, which increase the transmittance up to 95% – 97% for a given
wavelength range. Germanium and Silicon are used frequently for IR optical
components. Advances in IR technology have resulted in improvements in the
resolution of imaging systems and designers of imaging systems struggle to
develop matching optical systems. This has led to a search for achromatic IR
optical materials and the use of chalcogenide glasses. This selenium based
glasses combine well with other IR optical materials and provide
high-resolution optical material operating in the 3–5 and 8–12 ?m atmospheric
windows.

First and second generation
scanning systems used configurations of lenses, rotating prisms, rocking
mirrors or rotating multisided mirror drums. The precise design was dependent
on commercial features and the functional purpose of the imager. Single-element
detectors have the advantage of simplicity, both electronically and
mechanically. There are two advantages in replacing a single detector by a
multi-element array of n similar detectors. There is a decrease in noise
level since the signal increases in proportion to n whereas noise
increases in proportion to n1/2 and the higher
scan speeds that can be obtained with array detectors make these instruments
important for investigating rapid temperature changes. In a clinical context,
high resolution real time imaging allows a precise focusing on the skin surface
and continuous observation of thermal changes so that transient or dynamic
studies can be made on patients.

A significant development in the imaging
technology was the SPRITE detector. The SPRITE CMT detector performs the delay
and add functions within the element so that a single detector replaces a
linear array of detectors. In a conventional in-line array, the signal from
each IR-sensitive element is pre-amplified and then added to the signal that is
generated in the adjacent element. In the SPRITE, the individual elements are
replaced by a single IR strip mounted on a sapphire substrate. It requires only
one amplifier channel and has optimum gain at high speeds. An eight-element
SPRITE detector is equivalent in performance to an array of at least 64
discrete elements but requires far fewer connections. By arranging the
detectors in a stack, outputs can be stored in parallel in-line registers and
serially combined to a TV compatible display rate.

 

Focal-Plane Staring
Arrays

Since
1993 much of the earlier technology that employed single detectors, linear
arrays and SPRITE detector arrays has been replaced by the development of
staring-array detectors. These third generation cameras offer higher
temperature resolution images in real time. The absence of a scanning mechanism
means that all FPA solid state cameras are very compact and quiet in use. The
scene is viewed through a lens and when appropriate, a filter can be
incorporated to view IR above a specific wavelength. FPA arrays have been
constructed from a number of different materials including InSb, PtSi, CdHgTe
and InGaAs. Complex scanning systems are no longer required and the inherent
simplicity of the staring-array detector together with advances in micro-cooler
technology has resulted in the manufacture of very compact, high performance
staring-array systems.

In
a scanning system, the detector or each pixel of the detector only sees the
object for a very short time and this reduces the amount of energy collected.
In FPA systems, the scanned detector is replaced by an array of detector cells,
one for each pixel staring constantly at the object being imaged. To increase
the number of detectors inside the sensor vacuum Dewar, most quantum detector
arrays operate in photovoltaic mode. The detectors can be fabricated on
substrate as p–n junctions using integrated circuit techniques with very
high packing density. There are two ways of constructing FPA detectors. Monolithic
detectors are easier and cheaper to construct because both the IR-sensitive
material and the signal transmission paths are on the same layer. On the
contrary, in the case of hybrid FPAs, the detector is on one layer and the
signal and processing circuitry is on another layer. The advent of micron and
sub-micron silicon technology has led to the manufacture of complex signal conditioning
electronics and multiplexers integrated onto a silicon chip. This in turn is
incorporated directly behind the IR detector within the vacuum encapsulation.
The problem of non-uniformity of detector response in FPAs is addressed by
using digital signalling electronics and computing technology to match all
channels.

Uncooled
FPA detectors based on the principle of the bolometer have also been developed.
These devices consist of a sensitive area whose electrical resistance is strongly
dependent on its temperature. The absorption of the incident thermal radiation
changes the temperature of the sensitive area and the change in the measured
electrical resistance results in a signal proportional to that radiation. The
disadvantage of this type of thermal detector is that they react relatively
slowly compared to the response of photon detectors. However, such detectors
respond fast enough to work well in FPA systems where response requirements are
in millisecond range.

Staring-array
technology has also been applied to quantum-well-type IR photodetectors
(QWIPs). These devices are built to have a quantum well with only two energy
states, the ground state and the first excited state. The excited state is
arranged to be near the top of the well so that it can detect light photons. By
alternating layers of the well material such as GaAs and the potential barrier,
it is possible to control the characteristics of the QWIP so that it will
respond to a particular wavelength of radiation. Normally, QWIP detectors are
designed to detect radiation in the 8–9 ?m range.

FPA-based
systems requiring detector cooling have also been developed. In these cameras,
the matrix of detector cells is fashioned from InSb or from PtSi and must be
cooled to 80 K for optimal use as a thermal imaging device. PtSi has reliable
long-term stability but it has low quantum efficiency. It is sensitive to
radiation in the range of 1–5 ?m. Cooling of FPA detectors is usually
accomplished by a Stirling cooler. Typically, matrices used in clinical imaging
are either 320 × 240 pixels or 640 × 320 pixels but for research purposes,
detectors with arrays of 512 × 512 pixels and 1024 × 1024 pixels have been
developed.

The
image quality of FPA detectors that are used clinically is superior to that of the
previous scanning systems. Image capture and image processing are easier and
faster. Clearly, the use of an uncooled device for dynamic imaging of patients
in a ward or clinic is advantageous.

 

Pyroelectric Imaging
Systems

The
pyroelectric effect is exhibited by certain ferromagnetic crystals such as
barium titanate and triglycine sulphate (TGS). When exposed to a change in
radiance, these materials behave like capacitors on which electrical charge
appears. The magnitude of the effect depends on the rate of temperature change
in the detector, so the sensor does not respond to a steady flux of radiation.
Pyroelectric detection has been developed as a cheaper alternative to
photon-detector based systems. Although pyroelectric detectors were originally
incorporated into systems employing mechanical scanning devices to construct a
thermal image, most value for clinical work came from the development of
pyroelectric vidicon camera tubes. In pyroelectric systems the scene is panned
or modulated by a rotating disc and the IR radiation enters the vidicon tube by
means of a germanium IR transmitting lens (8–14 ?m) which focuses the image of
the thermal scene onto a thin disc of TGS pyroelectric material. At the front
of the TGS target, there is an electrically conducting layer of material that
is chosen to be a good absorber of thermal radiation. The target is scanned in
a TV raster by the electron beam of the vidicon tube and the image is displayed
on a TV monitor. The latest pyroelectric imagers are based upon FPAs that use
pyroelectric effect in ceramic barium-strontium titanate. Multi-pixel arrays
having an NETD of 0.5°C have been developed largely for industrial use and
surveillance purposes.

 

 

Applications

 

Rheumatoid arthritis

Rheumatoid
arthritis is a chronic inflammatory disorder that affects the joints. It results in overperfusion of the
tissue and a consequential increase in the skin temperature. Thermal imaging
provides objective, quantifiable and reproducible measures of the intensity and
the extent of joint involvement. It can be distinguished between deep-seated
inflammation and more cutaneous involvement. By standardising the conditions
and cooling the peripheral joints so that the skin is within a specific
temperature range (26°C–32°C for the lower limbs and 28°C–34°C for the joints
of upper limbs), it is possible to quantify the thermal pattern in the form of
a thermographic index on a scale from 1.0 to 6.0, in which healthy subjects are
generally found to be less than 2.5 while inflammatory joints rise to 6.0. This quantitative analysis is a very
effective means of assessing the efficacy of anti-inflammatory drugs used in
the treatment of rheumatic conditions.

 

Raynaud’s disease

Raynaud’s disease is a medical
condition in which the spasm of the fingers’ arteries cause episodes of reduced
blood flow, provoked by cold or emotional stress. Through thermal imaging, the
severity of the disease can be quantified and consecutive attacks can be
compared with each other. Due to the difference in the underlying disease
processes, the primary and secondary forms of Raynaud’s disease can be
differentiated by thermal imaging, which provides valuable information for
further diagnostic procedures and individualised management of the disease and
also has prognostic value.

 

Knee osteoarthritis

Knee osteoarthritis is the
degenerative disease of the tissues of the knee joint, accompanied by an
inflammatory process of varying degrees. Morphological changes observed by
imaging methods can be detected only after a long period of time from the onset
of signs and symptoms of the disease, even when sensitive imaging methods are
used. For these reasons, the evaluation of therapeutic efficiency and the
decision on the continuation of a certain therapy or its replacement by an
alternative therapy cannot be based on the results of the traditional imaging
modalities during the early modifiable course of the disease. The patella
physiologically represents a cool spot with a characteristic shape in the thermal
imaging, because its thick bone plate prevents the dissipation of the heat
produced by the knee joint through the patella and thus the heat is dissipated
around the patellar margin, which can be detected by the presence of a slightly
warmer band that surrounds the patella. In case of inflammatory processes of the
knee, the normally cool spot that represents the patella with the surrounding
slightly warmer band becomes distorted or disappears completely and the
temperature of the skin covering the inflamed knee tissues rises. Even in advanced
osteoarthritis that is detectable in X-rays, the increased temperature of the
skin covering the patella correlates with the severity of the radiographic
changes. Quantitative assessment of pain-related thermal dysfunction by thermal
imaging is utilized in other parts of the body as well.

 

Plastic and
reconstructive surgery

In plastic and reconstructive
surgery thermal imaging is used in many ways. It is an excellent diagnostic
tool to identify the dominant perforator vessels before free flap surgery,
which helps in pre-operative planning. It is an exceptional method to monitor
the perfusion of the free flap after connecting its vessels-arteries and veins
to the site of reconstruction intraoperatively. In the post-operative period it
is a sensitive, valuable method to assess the free flap in difficulty and to
decide if the clinical symptoms are related to problems with flapper fusion or
are due to other causes such as infection.

 

Analysis of cortical
cerebral perfusion by the cold saline technique

Minute changes of less than 0.01 K in
the temperature of the cerebral cortical surface can be detected by thermal
imaging. It has led to the proof of concept study of measuring the cortical
cerebral perfusion by the cold saline technique. A small amount -10mL of ice cold
saline was administered as a bolus into a central vein and subsequent changes
in cerebral cortical temperature have been recorded by infrared video
thermography and analysed by the principal component analysis (PCA) in patients
who were operated on for their cerebral pathologies such as ischemic stroke,
brain tumour etc. It has been shown, that the method is able to differentiate
between cortical regions with good or poor perfusion.

 

Infrared video
thermography as a touchless polygraph method for psychophysiological stress
detection

Infrared imaging and image analysis
have been introduced as a powerful tool for deception detection around 2000 by
the demonstration of its ability to detect facial stress patterns at a
distance. The breathing cycle can be monitored by thermal imaging based on the
difference in temperature between the exhaled air and the ambient temperature. The
cardiac pulse wave can also be monitored by infrared thermography in locations,
where large arteries travel close to the skin surface.  Replacement of the traditional polygraph
testing by non-contact infrared video thermograph recordings and their
multifaceted analysis makes it possible to test a large number of individuals
for potential signs of their deceptive behaviour.