Optical coherence tomography powerpoint presentation

Optical Coherence Tomography (OCT) is a pivotal advancement in the field of biomedical imaging, offering unparalleled resolution and depth in visualizing the internal structures of tissues. This non-invasive imaging modality leverages the principles of low-coherence interferometry to generate cross-sectional images of biological tissues. The technique, similar to ultrasound imaging but employing light rather than sound, measures the echo time delay and intensity of backscattered light from different layers within a tissue. This essay delves into the technical principles, applications, advancements, and future potential of OCT, with a focus on its use in ophthalmology, cardiology, dermatology, and oncology.

### Technical Principles of OCT

At the core of OCT technology is the concept of low-coherence interferometry. OCT uses a broadband light source, typically near-infrared light, which is split into two beams: a reference beam and a sample beam. The sample beam is directed at the tissue, and the light is scattered and reflected back from various depths within the tissue. The reference beam travels a fixed path and is then combined with the light returning from the sample. The interference pattern resulting from this combination is analyzed to determine the time delay and intensity of the backscattered light, which corresponds to different tissue depths.

The axial resolution of OCT is determined by the coherence length of the light source, with shorter wavelengths providing higher resolution. The lateral resolution depends on the focusing optics used in the system. The typical resolution of OCT ranges from 1 to 15 micrometers, allowing for the visualization of microstructures within tissues. The depth of penetration varies with the wavelength of light used; longer wavelengths can penetrate deeper into tissues, albeit with lower resolution.

### Applications of OCT

#### Ophthalmology

OCT has revolutionized the field of ophthalmology, becoming a standard tool for diagnosing and monitoring a wide range of ocular conditions. Its ability to provide high-resolution images of the retina and optic nerve head is invaluable in detecting and managing diseases such as age-related macular degeneration (AMD), diabetic retinopathy, and glaucoma.

1. **Macular Degeneration**: OCT can visualize the layers of the retina and detect the presence of drusen, which are deposits associated with AMD. It also helps in identifying fluid accumulation and neovascular membranes, which are critical for diagnosing and managing wet AMD.

2. **Diabetic Retinopathy**: In diabetic patients, OCT aids in the detection of macular edema, a common complication that can lead to vision loss. The technology helps in assessing the extent of retinal thickening and the response to treatments like anti-VEGF injections.

3. **Glaucoma**: OCT is instrumental in measuring the thickness of the retinal nerve fiber layer (RNFL), which is essential in the early detection and monitoring of glaucoma

1. University of Cyprus BiomedicalImagingandAppliedOptics Biomedical

Imaging

and
Applied

Optics
Optical Coherence Tomography

2. Optical Biopsy Definition • The in situimaging of tissue
itt ith lti
m
icros
t
ruc
t
ure w
ith
a reso
lu
ti
on
approaching that of histology, but
without the need for tissue
excision and processing
22

3. Optical Coherence Tomography
•Analogous to ultrasound,
except that it measures
intensity of back-reflected
li
g
ht
g
•Technologically different than
ultrasound
• Can achieve resolutions in the
order of 1-10 μm at depths of 2-
3 mm
• Contrast provided by index of
refraction mismatch in tissue
• Interferometric de
p
th
p
localization
33

4. Optical Coherence Tomography
1 mm
CT
MRI
CT
tion
100 μm
US
Resolut
10 μm
End
1 μm
Micr
CM
Sub-cellular
Resolution
Penetration
100 μm 1 mm 1 cm
10 cm 1 m
4 4
Penetration

5. History
•Michelson Interferometer
• End of 19
th
century
•Optical Coherence Domain
Reflectrometry (OCDR) Reflectrometry

(OCDR)
• Optical testing of electronics
(Youngquist 1987)
• Fault location in wave
g
uides
(
1987
)
g()
• Eye length measurement (Frecher
1988)
•
Optical Coherence Tomography
•
Optical

Coherence

Tomography

(OCT)
• First applied to transparent tissues
in ophthalmology (Fujimoto 1991) in

ophthalmology

(Fujimoto

1991)
• Subsequent development of
technology resulted in application in
scatterin
g
tissues
55
g
(Huang et al, Science, 254, 1178-1181, 1991)

6. OCT Principles
λ
Coherent Laser Source
Reference
ΔL
2
Beam
Splitter
ΔL
Sample
Source
λ
Low coherence Source
λ2
dz
Detector
ΔL
6 6

7. OCT Principles
Transverse Scanning
Backscatter Intensity Backscatter

Intensity
th) ing (Dept ial Scann Axi
Tissue Specimen
77

8. OCT Principles
•Monochromatic source •
Intensity at the detector
()
s
s
L
Et
c
−
()
r
r
L
Et
c
−
•
Intensity

at

the

detector
{
}
() 2Re () ( )
dsr sr
III EtEt
τ
τ
∗
=++ +
Bd t (l h )
τ
τπτ
=++
0
() 2 ()cos(2 )
dsrsrtc
IIIIIV f
0
2
() ()
if
tc
Vt Ate
π
τ
=
•
B
roa
d
spec
t
rum
(l
ow co
h
erence
)
source
2
() ()
x
x
jf
Sf re d
xx
πτ
τ
τ
∞
−
=
∫
*
() E () ( )
[]
x
x
rxtxt
τ
τ
=−
xx
−∞∫
{}
τπτ
∞
=ℑ = −
∫ 0
( ) ( ) ( )exp( 2 )
tc
VSkSfjfdf
88
0
{
}
Δ
=++ ℑ Δ
0
() 2 ()cos( )
dsrsr
ILII II Sk kL

9. OCT Principles •Axial resolution
λ
)
2
ln(
2
2
dz
oλ
λ
π
)
2
ln(
2
Δ
=
• Choice of source
(
λand Δλ
)

dz
(
)
affects axial resolution (dz) but
also the penetration (μs)
dx
9 9

10. OCT Principles •Transverse resolution
4
f
λ
⎛⎞
4
f
dx
d
λπ
⎛⎞
=
⎜⎟
⎝⎠
• Choice of focusin
g
o
p
tics
b
Low NA
gp
(NA=d/f) affects transverse
resolution (dx) and depth of focus
(b)
dx
1010

11. OCT Principles •Transverse resolution
4
f
λ
⎛⎞
4
f
dx
d
λπ
⎛⎞
=
⎜⎟
⎝⎠
• Choice of focusin
g
o
p
tics
b
High NA
gp
(NA=d/f) affects transverse
resolution (dx) and depth of focus
(b)
dx
1111

12. OCT Principles
Optical
Source
Reference
Arm
(NIR)
Beam
Splitter
Tissue
Display
Sample
Arm
Tissue
Electronics
Computer
Arm
y
x
1212
x

13. OCT System Design
Optical
Source
Reference
Arm
Scanning Schemes • Time Domain
• Fourier Domain
(NIR)
Beam
Splitter
Tissue
Display
Sample
Arm
Tissue
Electronics
Computer
Arm
y
x
1313
x

14. Time-Domain OCT
•Axial scanning by modifying the
Retroreflector
reference arm length in time
• Galvanometric mirror scanning
•< 1
00

A
-
Sca
n
s/sec

100

A
Scans/sec
• Simplest option but also the most
slow
•
Piezoelectric fiber stretching
Galvanometer
Piezoelectric

fiber

stretching
• < 400 A-Scans/sec
• Faster but introduces variable
dis
p
ersion which de
g
rades
pg
resolution
• Helical Rotating Mirror
• <4 000
A
-Scans/sec
Fiber Spool
• Very expensive to manufacture
• Proprietary
Fixed Mirror
1414
Piezoelectric

15. Time-Domain OCT
• Optical phase delay line
• The technique was originally
developed for femtosecond pulse
measurements
based on Fourier
transform pulse
Collimator
•
based

on

Fourier
-
transform

pulse

shaping techniques.
• Relies on the basic property of the
Fourier transform
θ
i
f
γ
Grating
Λ
θ
o
x(λ)
x
o
θ(λ)
(
)
(
)
{
}
00
exp
x
tt X jt
ω
ω
ℑ
−←⎯→−
• phase ramp in the Fourier domain
corresponds to a group delay in
the time domain.
l
Scanning
Mirror
Rotation
axis
• 2-4 000 A-Scans/sec
• More complicated but faster
l
axis
1515

16. Fourier-Domain OCT •Detect the spectrum and take the Fourier Transform to retrieve
the A-Scan
•Detecting individual wavelengths while keeping the reference
arm fixed
• Spectrograph
St
•
S
wep
t
source
•Advantages
20 000
200 000 A
S/
•
20

000

–
200

000

A
-
S
cans
/
sec
• Improved SNR
Di d t
•
Di
sa
d
van
t
ages
• More expensive hardware •
More demanding post
-
processing
1616
More

demanding

post
processing

17. Fourier-Domain OCT •Spectral FD-OCT
Optical
Source
Reference
Arm
(NIR)
St h
CCD
Beam
Splitter
Sample
S
pec
t
rograp
h
Display
Sample
Arm
Sample
Computer
Arm
1717

18. Fourier-Domain OCT •Swept Source OCT
Swept Optical Source
Reference
Arm
th Wavelengt
Ti
Beam
Splitter
Sample
W
Ti
me
Display
Sample
Arm
Sample
Digitizer
Computer
Arm
1818

19. Fourier-Domain OCT
Sample
Reference
Plane
Fourier
Transform
ntensity
Transform
In
WavelengthDistance
1919

20. Fourier-Domain OCT
Sample
Reference
Plane
Fourier
Transform
ntensity
Transform
In
WavelengthDistance
2020

21. Fourier-Domain OCT
Sample
Reference
Plane
Fourier
Transform
ntensity
Transform
In
WavelengthDistance
2121

22. Fourier-Domain OCT
(
)
(
)
2
exp( 2 ) ( )exp( 2 ( ( ). )
R
Ik Ska jkr az jkr nzzdz
∞
=++
∫
(
)
(
)
0
exp( 2 ) ( )exp( 2 ( ( ). )
R
Ik Ska jkr az jkr nzzdz
++
∫
k:wavenumber k=2π/λ r
:
path
length
in
the
reference
arm
r
:
path
length
in
the
reference
arm
r+z:path length in the object arm
z:path length in the object arm, measured from the reference plane
z
0
:offset distance between reference plane and object surface
fti
id
(
1
f
d
i
ddi
th
l
f
n:re
f
rac
ti
ve
i
n
d
ex
(
n=
1
f
o
r
z<z
0
an
d
vary
i
ng
d
epen
di
ng on
th
esamp
l
e
f
o
r
longitudinal positions in the object z > z
0
)
a
R
:reflection coefficient of the reference
a
(
z
)
:
b
ackscatterin
g
coefficient of the ob
j
ect si
g
nal
,
a
(
z
)
is zero fo
r
z<z
0
()
g
j
g,
()
0
S(k):spectral intensity distribution of the light source
μμ (2) (2)
1
() ()1 () [()]
time scale time scale
jknz jknz
Ik Sk â d ACâ d
∞∞
⎛⎞⎜⎟
∫∫
(2) (2)
() ()1 () [()]
e
4
11
jknz jknz
Ik Sk â
ze
d
z
AC â
z
d
z
−
−
−∞ −∞
⎜⎟
=
++
⎜⎟
⎝⎠
⎛⎞
∫∫
2222
(
)
{
}
()
{
}
11
() ()1
28
zz
Ik Sk âz ACâz
⎛⎞
=+ℑ+ℑ⎡⎤
⎜⎟⎣⎦
⎝⎠

23. Fourier-Domain OCT
Period of cosine frin
g
es
{} {} ()
()
()
(
)
11
11
() ()
28
FIk FSk z âz ACâz
ABCD
δ
−−
⎛⎞
=⊗++
⎡
⎤⎡⎤
⎜⎟
⎣
⎦⎣⎦
⎝⎠
⊗++
g
max
dk
nz
π
=
(
)

ABCD
=⊗++
A
BAD
⊗
+⊗
Maximum resolvable depth
2
λ
0
max
4( )
Z
ndλ
λ
=
.dλ=wavelength sampling interval (defined
by
the
detector
separation
(defined
by
the
detector
separation
of a linear detector array) Sensitivit
y
Gain
A
C
⊗0z
0
z
y
22
s
hot shot shotFWHMRD
FD TD TD
Z
SSS
ddz
λ
λ
Δ
=≈
2323

24. OCT System Design
Light sources
Optical
Source
Reference
Arm
• Superluminescent diodes
• Semiconductor amplifiers
• Femtosecond lasers
St
(NIR)
•
S
wep
t
sources
•…
Beam
Splitter
Tissue
Display
Sample
Arm
Tissue
Electronics
Computer
2424

25. Light Sources For OCT
•Choosing the light
Dominating Loss in OCT
source
• Four primary
considerations
S
• wavelength,
• bandwidth,
•
power (in a single
-
cattering
power

(in

a

single
transverse-mode),
• stability (portability,
ease-of-use, etc)
Coefficien
Scatterin
g
nt (rel)
g
2525

26. Light Sources For OCT •Light source spectrum
• Basic property
• the temporal coherence envelope
function G(τ) is related to the
power spectral function S(
ν
)
through
G(τ) = FT{S(ν)}
•
Wiener
Kinchine theorem
•
Wiener
-
Kinchine

theorem
• broadband source ⇔high axial
resolution
FFT
dz
o
λ
)2ln( 2
2
=
dz
λ
π
Δ
=
26 26

27. Light Sources For OCT •Light source spectrum
• Basic property
• the temporal coherence envelope
function G(τ) is related to the
power spectral function S(
ν
)
through
G(τ) = FT{S(ν)}
•
Wiener
Kinchine theorem
•
Wiener
-
Kinchine

theorem
• broadband source ⇔high axial
resolution
dz
o
λ
)2ln( 2
2
=
dz
λ
π
Δ
=
27 27

28. Light Sources For OCT •Continuous sources
• SLD/LED/superfluorescent fibers,
• center wavelength;
• 800 nm (SLD), 1300 nm (SLD, LED), 1550 nm, (LED, fiber),
• power: 1 to 10 mW (c.w.) is sufficient,
• coherence length;
• 10 to 15 μm (typically),
•Pulsed lasers
• mode-locked Ti:Al2O3 (800 nm), •
3 micron axial resolution (or less)
•
3

micron

axial

resolution

(or

less)
.
•Scanning sources
• tune narrow-width wavelength over entire spectrum, • resolution similar to other sources,
• Fourier Domain OCT
• advanta
g
e that fast scannin
g
is feasible.
2828
gg

29. OCT System Design
Optical
Source
Reference
Arm
(NIR)
Beam
Splitter
Tissue
Bdli db
Display
Sample
Arm
Tissue
B
eam
d
e
li
very an
d
pro
b
es
• Ophthalmoscope
• Catheter
Hand
held probe
Electronics
Computer
Arm
•
Hand
-
held

probe
• Microscope
2929

30. Imaging Devices •Application dependent •Ophthalmoscope
• Most widely used OCT instrument • Time-domain systems:
• Zeiss Meditec
• Fourier-domain systems
• Zeiss Meditec
• Heidleberg Engineering
•O
p
tovue
p
• Topcon
• Some combine OCT with
scanning laser ophthalmoscopy scanning

laser

ophthalmoscopy
Stratus™ HD-OCT system from Carl
Zeiss Meditec
3030
Zeiss

Meditec

31. Imaging Devices •Catheters
• Diameter < 1 mm
• Even as small as a needle
•
Scanning Schemes Scanning

Schemes
• Push-Pull
• Rotational
•
Helical (volumetric) Helical

(volumetric)
• Availability
• LightLab, Inc., Helios balloon
catheter catheter
• Custom catheters
3131

32. Imaging Devices •Handheld Probes and
Microscopes
• Usually employ orthogonal
g
alvanometricall
y
scanned
gy mirrors
• Dermatologic, biological and
research a
pp
lications
pp
• Availability
• Thorlabs •
Bioptigen
•
Bioptigen
•Custom
3232

33. OCT System Design
Optical
Source
Reference
Arm
(NIR)
Beam
Splitter
Tissue
Computer control • Drive system
• Real-time display
Display
Sample
Arm
Tissue
• Data management Image & signal processing •
Motion reduction
Electronics
Computer
Arm
•
Motion

reduction
• Speckle reduction • Image enhancement •
Rendering algorithms
3333
Rendering

algorithms
•…

34. Functional OCT •Doppler OCT
• Motion of scatterers imparts a
frequency shift in the OCT signal
λ
λ
=±
00
22
ref sc
d
vv
f
• Detection
• Frequency analysis of OCT
ifinter
f
erogram, o
r
• Phase sensitive imaging
• Can detect micro-flows
3434

35. Functional OCT •Doppler OCT
35 35
Retinal blood vessel flow
Letigeb et al, Opt. Express 11, 3116-3121 (2003)

36. Functional OCT •Constructive interference
• Phase matched
• Polarization matched •Birefringence
• One polarization retarded
more than other more

than

other
•Polarization Sensitive OCT
• Interference can be detected
separately for orthogonal
polarizations
•De
p
th-resolved Stokes
p
parameters can be calculated
• Useful for birefringent materials
•
i e collagen layers
3636
i.
e
.
collagen

layers

37. Functional OCT
3737
Birefringence of bovine muscle before, during and after laser exposure deBoer et al, Opt. Lett. 22, 1439-1441 (1997)

38. Functional OCT •Spectroscopic OCT
source
detector
Amplitude (envelope) of the OCT signal
FFT
Interferometric OCT signal
λ
Standard OCT Demodulation - Envelope
Spectroscopic OCT Interferometric Signal ÆSpectral
⇒Reflectivity / Scattering
differences (centroid, skewness, shape)
⇒Spectrally dependent scattering
Ab ti
38 38
⇒
Ab
sorp
ti
on

39. Functional OCT
Normal Esophagus Barrett’s Esophagus Normal Cervix
•Lon
g
er wavelen
g
ths
p
enetrate dee
p
e
r
39 39
ggp p
•Areas of wavelength-dependent backscattering are visible

40. Functional OCT •Mechanical Properties of Tissue
• Displacement and strain maps
from changing forces
OCT
Elastography
•
OCT

Elastography
• Elasticity measurements can be
made from slight mechanical
df ti f
ti
d
e
f
orma
ti
ons o
f

ti
ssue
• Correlation of successive images
• Phase-resolved imaging
• Depth-resolved, high resolution
4040

41. Functional OCT
4141
OCE of breast tissue
Liang et al, Optics Express,
16:11052-11065, 2008

42. Software and Algorithm Issues in OCT •OCT is now a high-data-rate
streaming technology
• Example: to cover 7x20 mm of
area in the eso
p
ha
g
us re
q
uires
pg q
• 1400 AScans x 1000 pix/Ascan x
14 bit = 78.4T bits
• Sampled at 20 MS/s = 280 Mbit/s
•Requires very demanding post
processing
• For real-time display: 20 000
times per second
• FFT
• Filter
•Scale
• Color code
4242

43. Software and Algorithm Issues in OCT •Issues still unresolved
• Optimal filtering
• Exponential decay correction
•
Segmentation
•
Segmentation
• Display • Visualization
4343

44. Applications
•OCT can play a role in early diagnosis of
disease and improve patient prognosis
•High resolution imaging :
Sifdi hbii
•
S
creen
ing
f
or
di
sease w
h
ere
bi
opsy
is
impossible, difficult or hazardous
• Guiding biopsies to improve sensitivity and
specificity and reduce the number required
• Non-invasive monitoring of response to
therapy •Recent developments
• Increased speed (up to 200 fps) • Improved resolution (1-5 μm)
• Compact and reliable systems
4444

45. Ophthalmology •Diagnosis and management of
ldi d
ocu
l
ar
di
sor
d
ers
• Age-related macular degeneration
• Diabetic macular edema
• Macular hole • Epiretinal membrane • Glaucoma
•Latest developments
• Combining OCT with fundus
photography and scanning laser photography

and

scanning

laser

ophthalmoscopy
• 3D visualization of tissue
morphology morphology
• Ultra-high speed, ultra-high
resolution OCT with adaptive
o
p
tics and
p
ancorrection
Standard (10 μm) and UHR (2 μm) OCT of the retina Drexler, Fujimoto, Progress in Retinal and Eye Research 27 (2008) 45
88
45 45
pp
Research

27

(2008)

45
–
88

46. Ophthalmology
Macular Hole Drexler Fujimoto Progress in Retinal and Eye Research 27 (2008) 45
–
88
4646
Drexler
,
Fujimoto
,
Progress

in

Retinal

and

Eye

Research

27

(2008)

45
–
88

47. Ophthalmology
3D UHR Ima
g
in
g
of the retina
gg
with OCT fundus view Drexler, Fujimoto, Progress in
Retinal and Eye Research 27
(2008) 45
–
88
4747
(2008)

45
88

48. Ophthalmology
Imaging of individual photoreceptors Drexler, Fujimoto, Progress in Retinal and Eye
Research 27 (2008) 45–88
4848

49. Cardiology •Visualization of the vessel wall
at the microscopic level
• High resolution imaging of
coronar
y
architecture
y
• Precise characterization of plaque
architecture
•
Quantification of macrophages
•
Quantification

of

macrophages

within the plaque
•Identification of the most
common type of vulnerable
plaque, the thin-cap
fibroatheroma
•Monitoring stent deployment
Lipid rich plaque (OCT and IVUS) and
49 49
Lipid

rich

plaque

(OCT

and

IVUS)

and

Macrophage content Low et al, Nature Clinical Practice, 3, 145-162, 2006

50. Cardiology
Stented Coronary Artery Yun
,
et al
,
Nature Medicine
,
12
,
1429-31
,
2006
5050
,, ,,
,

51. Gastroenterology
•Especially relevant application
for OCT
• High incidence
• Clinical benefits of early detection
• Need for pre- and post-treatment
assessment
•OCT a
pp
lied to the GI
pp
• Early detection of cancer in
Barrett’s esophagus patients
•Stud
y
of inflammator
y
bowel
yy
diseases in the colon
•Major goal = guidance of
excisional biopsy excisional

biopsy
•High resolution and high speed
now allow whole organ
i
Barett;s without and with HGD Image Size: 2.5 mm, Resolution: 10 x 20 μm
Evans et al, Clin Gastro Hepatol, 4, 38-43, 2006
5151
survey
i
ng

52. Gastroenterology
Volume:
7 x 20 x 1.6 mm
Resolution: 6 x 8 μm)
Scan Speed:
62k A
S/
Normal Colon
62k
A
-
S
cans
/
sec
5252
Normal

Colon
Adler, et al, Optics Express, 17(2), 784-796, 2009.

53. Gastroenterology
Volume:
7 x 20 x 1.6 mm
Resolution: 6 x 8 μm)
Scan Speed:
62k A
S/
Ulcerative Colitis
62k
A
-
S
cans
/
sec
5353
Ulcerative

Colitis
Adler, et al, Optics Express, 17(2), 784-796, 2009.

54. Dermatology •OCT can image
• The stratum corneum of
g
labrous skin
g
(palmoplantar)
• The epidermis and the upper dermis
• Skin appendages and blood vessels
•Uses
Ufli
ii itift
•
U
se
f
u
l
in non-
invas
ive mon
it
or
ing o
f
cu
t
aneous
inflammation, hyperkeratotic conditions and
photoadaptive processes
• May be of great value, in particular in cosmetics
and the pharmaceutical industry
•
Could potentially allow the differentiation between
•
Could

potentially

allow

the

differentiation

between

benign and malignant tissues. Beyond a high
resolution morphology in OCT images, tissue
•Additional parameters
• Scattering coefficient • Refractive index.
• OCT spectroscopy
• Tissue birefringence
• OCT elastography
•
OCT Doppler flow OCT

Doppler

flow

•Significant new insights in skin physiology
and pathology
Normal finger skin (standard OCT and
f)
5454
bire
f
ringes
)
Pierce, et al, J Invest Dermatol 123:458 –463, 2004

55. Dermatology
Nldfi ki N
orma
l
an
d
scar
fi
nger s
ki
n
(standard OCT and birefringes) Pierce, et al, J Invest Dermatol
5555
123:458 –463, 2004

56. Dentistry •Potential applications in
dentistry
• Detection of hidden dentinal
caries
• Quantitative monitoring of de- and
remineralisation of demineralised
lesions • Visualisation of interproximal
surfaces of premolars and molars
•
Investigation of the effectiveness
•
Investigation

of

the

effectiveness

of restorative fillings
• Early detection of soft tissue
diseases diseases
.
Ml t ti
5656
M
o
l
ar res
t
ora
ti
on
Brandemburg, et al, Optics Communications 227
(2003) 203–211

57. Developmental Biology •Imaging embryonic morphology
Ad t i th fil d f ti
•
Ad
vancemen
t

in
th
e
fil
e
d
o
f
gene
ti
cs:
genes identified, function studied
• New animal models to study gene
ex
p
ression or lack of ex
p
ression
pp
(knockout animals)
• Study of developmental and
embryologic changes require
termination termination
• High resolution imaging for
developmental biology has the
potential for non-invasive:
• Image embryonic microstructure and
phenotypic expression
• Image function and response to
medications
• Guide interventions and fetal
manipulation
• Monitor response to therapy
Xenopus Laevis morphology
Boppart, et al, DEVELOPMENTAL BIOLOGY 177 54
63 (1996)
5757
177
,
54
–
63

(1996)

58. Developmental Biology
•
Anterior eye: cornea
•
Anterior

eye:

cornea
,
lens, and iris.
•Corneal thickness ~ 10
μm
•Posterior eye: ganglion
cell layer retinal cell

layer
,
retinal

neuroblasts, and
choroid
Xenopus Laevis morphology Boppart, et al, DEVELOPMENTAL BIOLOGY 177 54
63 (1996)
5858
177
,
54
–
63

(1996)

59. Developmental Biology
Developing Embryo
5959

60. Developmental Biology
Stage: 4 cell, 1 hour Stage: 32 cell, 1.75 hour Stage: Prim-5, 24 hr Stage: Hatched, 48 hr
Developing Zebra fish embryo Boppart, et al, DEVELOPMENTAL DEVELOPMENTAL

BIOLOGY 177, 54–63
(1996)
6060

61. Developmental Biology
Abnormalities in Xeno
p
us Laevis mor
p
holo
gy
6161
ppgy
Boppart, et al, DEVELOPMENTAL BIOLOGY 177, 54–63 (1996)

62. Developmental Biology
4D imaging of mouse embryonic heart J ki t l O ti E 14 736
748 (2006)
6262
J
en
ki
ns, e
t
a
l,
O
p
ti
cs
E
xpress,
14
,
736
-
748
,
(2006)

63. Commercial OCT Systems •First commercial OCT devices
• Humphrey Systems (now a part
Carl Zeiss Meditec, Inc.)
•
retinal imaging retinal

imaging
• released in 1996 • FDA approval in 2002.
St t OCT™ t lli
•
St
ra
t
us
OCT™
sys
t
ems se
lli
ng
more than 6000 units
• Cirrus™ HD-OCT system
6363

64. Commercial OCT Systems •Many more devices available now
•
LightLab Imaging
•
LightLab

Imaging
• Imalux Corporation,
• ISIS Optronics GmbH
• OCT Medical Imaging, Inc
•
Michelson Diagnostics Ltd Michelson

Diagnostics
,
Ltd

• Novacam Technologies, Inc
• Lantis Laser Inc
• OptoVue, Inc
•
Topcon Corporation Topcon

Corporation
• Optol Technology • Heidelberg Engineering • Opthalmic Technologies, Inc • Thorlabs, Inc • Bioptigen, Inc
•Individual OCT components
• Femtolasers Produktions
• Nippon Telegraph and Telephone
Corporation
• Thorlabs, Inc
•
MenloSystems GmbH
6464
•
MenloSystems

GmbH

65. Can we “see” more?
•Resolution limit
• 1-10 μm
• Especially limited in the lateral
direction direction
•Many cancerous/pre-cancerous
changes in the μm range
• Cell proliferation: 5 μm spacing
change
• Nucleus variations: ~2-4
μ
m
μ
diameter change
• Sub-cellular and Sub-nuclear
variations:
<
1
μ
m change
variations:

1

μ
m

change
6565

66. Speckle Analysis
•Uresolvable features
Barett’s Esophagus
• Coherent technology Æspeckle
• Mostly treated as noise
Skl
•
S
pec
kl
e
• Contains information regarding the
•Size • Concentration
• Spacing
• Periodicit
y
100 µm
Dysplastic Cervix
y
•Etc • size and distribution of scatterers
•
Diagnostically useful information Diagnostically

useful

information
• Statistical and spectral properties
6666
100 μm

67. Speckle
•Signal scattered from a
0.8
1
Scatterer Distribution
distribution of scatterers
•Incoherent Scattering
•
Intensity summed
0.20.40.60.8
Intensity

summed
•Coherent scattering
• Field summed
100
200
300
400
500
600
700
800
900
1000
0
0.050.06
Incoherent Scattering
• Resulting intensity profile may not
resemble scatterer distribution
• Speckle
0
0.010.020.030.04
•Unresolvable Signal
• Contains information regarding the
number and distribution of
0.040.06
Coherent Scattering
100
200
300
400
500
600
700
800
900
1000
0
scatterers
• Diagnostically useful information
• Statistical and spectral properties
-0.04-0.02
0
0.02
6767
100
200
300
400
500
600
700
800
900
1000
-0.06

68. Speckle Analysis •Speckle properties reflected in
0.8
1
1μm
the spectral content
• Optical spectrum of OCT
•
S(
⎤
)
=
FFT{s(t)}
0.40.6
2μm 4μm
atter Intensity
S(
⎤
)

FFT{s(t)}
•S
p
ectral
p
ro
p
erties of
0
0.20.4
Backsca
ppp
backreflected signal depend on
• Source spectrum •
Effect of scatterers
1.22 1.23 1.24 1.25 1.26 1.27 1.28 1.29 1.3 0
Wavelength (nm)
•
Effect

of

scatterers
•Size ÆSpectral dependence of
back-scattered light (similar to
LSS
))
6868

69. Speckle Analysis
5000
•Solid Tissue Phantoms
3500400045005000
• Polystyrene sphere solutions in a
gel of acrylamide.
•
Diameter 1
μ
m,
2
μ
m and
4
μ
m
200025003000
Diameter

1
μ
m,

2
μ
m

and

4
μ
m
•[C]
1μm
=50 spheres, [C]
2μm
=5
spheres and [C]
4μm
=2 spheres
20
40
60
80
100
120
140
160
180
200
500
10001500
•Data acquired with TD OCT
• Imaging Volume: 17μm x 17μm x
15μm
20
40
60
80
100
120
140
160
180
200
• Collect images
• Digitized at 6x carrier frequency
•
Processed off
line
Image acquired from a
phantom of 1μm and [C]=50
•
Processed

off
-
line
69 69

70. Speckle Analysis
•Pre-
p
rocessin
g
5000
pg
• Normalization to RMS value of 1
•
Spectralestimation
3500400045005000
Spectral

estimation
• Autoregressive Power Spectral
Estimation
•
Effect of distribution and/or
200025003000
•
Effect

of

distribution

and/or

concentration
• Can be reduced by averaging of
spatially adjacent scans
20
40
60
80
100
120
140
160
180
200
500
10001500
spatially

adjacent

scans

• Spectra only affected by size
•
Classification
Image acquired from a
phantom of 1μm and [C]=50
20
40
60
80
100
120
140
160
180
200
Classification
•K-Means clustering
70 70
•Scatterer size estimation

71. Speckle Analysis •Autoregressive Power Spectral
Estimation
• Faster convergence from shorter
signals
•Burg’s method
• Minimizing (least squares) the forward
and backward prediction errors
• Number of coefficients: 100
• Size of window: 3001 (axial) x 25
(transverse) pixels
•Principal Component Analysis
(PCA)
•
Reduction of the number of variables Reduction

of

the

number

of

variables

(35)
• New orthogonal basis with no
redundant information
7171

72. Speckle Analysis •Classification
Mlti it A l i fV i
•
M
u
lti
var
ia
t
e
A
na
lys
is o
f

V
ar
iance
(MANOVA)
• New linear combinations of variables
• Maximum separation between
categories categories

• Discriminant based classification
•
Scatterer Size Estimation
•
Scatterer

Size

Estimation
• Principal Component Analysis (PCA)
• Solution to set of linear equations
j
i
i
xx
PA d= i
i
i
1
xxtraining
APd
−
=
7272

73. Speckle Analysis •Information can be extracted
from spectrum
• Classification of images based on
scatterer size
Intensity Image of Microspheres
Embedded in Acrylamide
• Sensitivity and specificity: 85-99
%
• Scatterer size estimation from
d=1
μ
md=2
μ
md=4
μ
m
solutions of linear equations
• Mean error: 16.5%
d=1

μ
m

d=2

μ
m

d=4
μ
m
Classification Results Overlayed on
Intensit
y
Ima
g
e
• Require a priori information and
training
yg
1 x 3 mm (17 x 20 μm)
73 73

74. Speckle Analysis •K-Means clustering •Divide the data into a predefined
number of clusters (groups)
•
Minimize the distance from the Minimize

the

distance

from

the

centroid of each group
• Assumes each object’s attributes
are coordinates in
multidimensional space
• Iteratively finds centroids and
reassigning the clusters to
minimize:
2
k
Vx
μ
=−
∑∑
• Advantage: No a-priori
information required
1
ji
ji
ixS
Vx
μ
=∈
=−
∑∑
7474

75. Speckle Analysis •Clusterin
g
areas of same intensit
y
but different size scatterers
gy
Intensity image of microspheres
Scatterer
Diam
1 μm2 μm4 μm
Intensity

image

of

microspheres
Diam
.
Sensitivity95.11 97.11 92.56
Sifiit
99 44
99 44
99 44
d=1 μm d=2 μm d=4μm
Clustering results of microspheres
S
pec
ifi
c
it
y
99
.
44
99
.
44
99
.
44
1 x 3 mm (15 x 30 μm)
75 75

76. Speckle Analysis
Intensity image of tadpole
v
n
Intensity image of nerve
v
f
Clustering image of tadpole
Clustering image of nerve
7676
1.5 x 1.95 mm (15 x 30 μm)
1.5 x 1.95 mm (15 x 30 μm)

77. Conclusions •OCT has shown great promise as a diagnostic tool •
UHRand ultra high speed systems can provide real time
•
UHR

and

ultra

high

speed

systems

can

provide

real

time
,
dynamic and diagnostically relevant information
•
There is still a lot to be done to confirm benefit to the patients
•
There

is

still

a

lot

to

be

done

to

confirm

benefit

to

the

patients
• Perform clinical studies
• Form standards and consensus
• Prove benefit to patients
7777