Available via license: CC BY-NC-SA
Content may be subject to copyright.
Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1 45
Access this article online
Website:
www.jcor.in
DOI:
10.4103/2320-3897.149379
Quick Response Code:
SamarthaClinicMumbai, 1P.D.HindujaNationalHospitalandMedical
ResearchCentre,Mahim,Mumbai,Maharashtra,India
Address for correspondence:Dr.B.K.Nayak,P.D.HindujaNational
HospitalandMedicalResearchCentre,VeerSavarkarMarg,Mahim,
Mumbai-400016,Maharashtra,India.E-mail:drbknayak@gmail.com
Manuscript received:08.12.2014;Revision accepted: 26.12.2014
Topography is the study of the shape of the corneal surface.
The early devices were limited by their measurements to
the central part of the cornea; however, with the explosion
of refractive surgical procedures on the cornea and its
consequences, the need to know more has led to newer devices
and technologies emerging in the market. With the pressures of
precise outcomes/results and the plethora of devices available,
it can be quite a problem to decide which technology to adopt
in clinical practice.
The investigative modalities for studying the corneal
shape have undergone a drastic make over in the past few
years. The once standard corneal videokeratoscopy has
now the company of the scanning slit, optical coherence
tomography (OCT), and Scheimpflug imaging to supplement
in the assessment of the corneal shape. The addition of
these devices has led to the addition of the word “corneal
tomography” in the medical jargon as far as corneal imaging
is concerned. This is because the images obtained by these
devices are essentially a cross section of the cornea, and the
elevation data thus obtained being analyzed further. It is in
contrast to enface images of concentric rings of the placido-
based devices.
Corneal refractive surgery in any form; surface treatment or
laser-assisted in situ keratomileusis (LASIK) leads to weakening
of the biomechanical strength of the cornea. This especially
when dealing with thinner and steeper[1,2] Indian corneas need
great attention to avert the eventuality of a post-refractive
ectasia later in patient’s lifetime.
The understanding of corneal topography and tomography
along with the advantages/disadvantages of each is
fundamental to the assessment of this risk before posting the
candidate for refractive surgery. The existing techniques have
evolved, new techniques have been added in the last few years.
The various techniques give us different information with a
different methodology, and it is important that we use the
best of each type to give the desired result in our procedures.
In the following article, we would illustrate the use
of these corneal imaging techniques in relation to their
importance in the pre-refractive surgery screening. The article
is not meant to be an exhaustive in content but will seek to
look at the topic from the point of view of a comprehensive
ophthalmologist wanting to take to learning the art of
interpretation of topography maps and using them for
decision-making. We also intend to touch upon the basics
and bring out certain technical aspects about the available
devices citing literature wherever possible, to let the reader
make an informed choice about the device (s) he/she would
like to use in his/her practice.
Currently, the corneal imaging techniques before refractive
surgery involve four main types of devices:
1. The videokeratoscope or the Placido-based devices, e. g.,
Topographic Modeling System (TMS) 4, Keratron, Atlas.
2. Scanning slit devices. E. g., Orbscan IIz.
3. The Scheimpflug devices. E. g., Pentacam, Sirius, and the
Galilei. The latter two have and additional large cone Placido
disc incorporated in them.
4. OCT-based devices (Visante from Zeiss).
The OCT will not be discussed in this article.
Commissioned Article
Corneal topography and tomography
Sachin Dharwadkar, B. K. Nayak1
Devices that evaluate corneal properties are an indispensible tool in a eye clinic nowadays. With the arrival of new
technology in addition to placido based devices, the options available now are many. Cornea based refractive surgery in
Indian eyes poses a challenge due to relatively thinner corneas. This is also compounded by lack of well dened, rigid
and universal criteria for case selection for the same. In this article we attempt to look at the most common methods
of corneal assessment in relation to the selection of candidates for corneal refractive surgery with a review of relevant
literature. This is not meant to be exhaustive, but a primer to ease the clinician into understanding and taking up to
learn and practice corneal evaluation.
Key words: Tomography, topography, corneal vertex, corneal apex, saggital curvature, elevation, reference surface
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
46 Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1
There are certain fundamental differences in the
videokeratoscopes, scanning slit, and the Scheimpflug devices
and the fact that their data is non-interchangeable has to
be kept in mind. They work on totally different principles;
have different methods of data acquisition, presentation,
and analysis. That apart, the same class of devices from
different manufacturers may not strictly compare.[3] All these
devices fundamentally involve simple principles of physics
combined with high-end mathematics and statistics (the
so-called artificial intelligence or neural networks), to give us
a simplistic view of the status of the patient’s cornea.
All the devices that are available can be considered to
consist of three main parts:
1. A projection device (back illuminated Placido rings, Blue
light emitting diode (LED)).
2. An acquisition device (a Charge coupled device (CCD)
camera for Videokeratoscope and scanning slit systems/
Scheimpflug camera for Scheimpflug devices), Spectrometer
for spectral domain OCT and.
3. An analytical device that is a computer with various
software (Neural networks/artificial intelligence) and
normative database (not in all the machines) to analyze
the data that is obtained above.
Let us now consider these techniques and their advantages
and drawbacks. The Penatacam printouts would be considered
as the prototype Scheimpflug device for the sake of
explanation. All other Scheimpflug-based devices offer similar
data but differ in their software (machine classifiers) and final
data outputs.
The basis of all topography and measurement/quantification
of the corneal surface started with the keratometer the
optical principle of which is depicted in Figure 1. Let us now
analyze the transition from these starts into the world of
videokeratoscopy.
Concept of Placido-Based Devices
Curvature measures bending. The more curved the surface the
smaller is the radius of curvature and higher the refractive
power. It is however important to remember that the radius of
curvature is an inherent property of a curved surface, whereas
the power [as shown in Figure 2] is derived from it with due
consideration to the refractive index of the media involved.
Also, different shapes can have the same power and hence
using the power as the surrogate for shape may not be an ideal
situation. The measurement of curvature of a given surface
can be done in various methods and with the help of various
devices. The keratometer is the reflection of the measurement
of the central cornea only, the mid-periphery requiring
instruments like the videokeratoscopes. Videokeratoscopes
cover the central 7-8-mm zone of the cornea.
The Placido disc invented by Antonio Placido in the late
1800s was the first attempt to qualitatively assess the shape
of the entire cornea. This consisted of a disc with concentric
dark and light rings in the center of which was a convex lens
for visualization of its reflection on the cornea. [Figure 3] The
disc in this case would be illuminated by an indirect source and
the observer would see the reflections of the rings or “mires”
on the examined cornea. The placement and the size of the
grid would indicate the kind of corneal problem qualitatively
as shown simplistically in Figure 4.
In the later part of the twentieth century, the devices
were developed that used a backlit Placido disc and a camera
respectively to image the cornea, instead of the indirect
illumination and the observers eye, so-called Photokeratoscopy.
This again was a qualitative test till sophisticated mathematical
software allowed the quantification of the curvature and the
shape and as a result of which the videokeratoscopy was born.
Corneal Videokeratoscopy (CVK)
This is the technique that has now evolved from the initial
efforts that started from the Placido disc devised by Antonio
Placido and have culminated into a high end gadget with
analytical tools. It essentially uses the reflection principle
and studies the first Purkinje image from the cornea resulting
from the reflection of the illuminated mires of the projection
device and its processing by the computer. The pre-corneal tear
Figure 1:PrinciplesofKeratometry(shouldbescannedfromthebook
page37(ABistheobjectandA’B’istheimage.Bymeasuringthesizeof
theobjectandimage,curvatureoftheconvexsurfacecanbecalculated
Figure 2:Asmallradiuscirclehasalargecurvatureandviceversa.
However,power cannottell shape,differentshapescanhave same
power
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1 47
film being the anterior-most layer of the cornea that reflects
the light, its nature has a great effect on the quality of the
images that are obtained for analysis. Some devices have also
introduced blue-colored mires (Nidek OPD scan 3) to facilitate
accurate edge detection by the computer.
The devices are of two types consisting of a large or a
small cone Placido projection system [Figure 5]. The large
cone Placido can be used slightly away from the patient’s
face, whereas the small cone device needs to be very near
the patient’s eye. The advantage of the small cone devices
is more complete coverage of the cornea and avoidance of
data loss due to the varying nasal bridge anatomy. These
areas of absent data are seen as data gaps or hatchings on
the final printout in most devices. In some, however, data
interpolation is used to give a complete coverage. It would be
worthwhile to see the acquired raw image before interpreting
the curvature data if not acquiring/selecting a suitable image
yourself. One must be aware of this fact, otherwise reports
can be misleading [Figure 6]. This probably is also the most
important disadvantage of the large cone Placido devices. The
theoretical disadvantages of the small cone devices however
is that they are more prone to focussing errors, the cone being
very near the patient’s eye. In clinical use, with modified cone
designs and software compensation, the small and large cone
devices are pretty much equivalent provided the area of cornea
mapped is almost equal. Seventy to ninety percent is the usual
coverage in large cone devices. The area covered is mentioned
on the printout for most machines.
The measurements of the two types of devices also differ
a little bit and the variation should be kept in mind in case
one shifts from one type of Placido device to another. Small
mire Placido devices represent the maximum curvature in
keratoconus at greater magnitude than large mire devices.[3]
The upper lid eyelashes however interfere with acquisition the
data in upper part of the cornea both these devices.
Corneal videokeratoscopy (CVK) measures central and
mid-peripheral corneal zones as opposed to conventional
keratometry that measures the central cornea and is especially
useful for evaluating irregular astigmatism compared to
the keratometer. Besides screening of refractive surgery
candidates, the other applications of CVK include: Diagnosis
of corneal irregularities (moderate to advanced ectasias,
dystrophies, surface disease, contact lens (CL) warpage,
scars, degenerations), evaluating unexplained visual loss,
Figure 3:Thisistheschematicdiagramof thePlacido’sdisc.There
areconcentricblackandwhiteringswithaconvexlensinthecenter
apertureasshown
Figure 4:Theimages in the toprowshow normal variations in the
cornealshape.Theimagebelowshowsinferiorsteepening(miresare
closelyplacedinferiorly)asseenabnormalcorneas
Figure 5: This gure shows the different cone designs in Placido
systems.The gureonthe leftside showsalarge conedeviceand
theoneontherightshowsasmallconedevicewherethemiresare
closelyarrangedandtheconeissmallerinsize
Figure 6:LargeconeandsmallconePlacidomiresinsameeye.Note
thedifferenceinthenumberofmiresandthedatalossduetonasal
anatomyin large conePlacido. Also notethatthe smallconemires
havelessdistinctedges
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
48 Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1
management of surgical patients (planning and monitoring
corneal grafts, refractive procedures, cataracts, pterygia), and
contact lens fitting.
The data obtained by the CVK is presented usually
as four topographic display formats: They are namely
curvature (axial and instantaneous/tangential), power
(refractive), and recently elevation (difference or relative
maps have been added for the analysis of anterior surface),
and wavefront maps have also been added. Elevation data,
however, when generated from a Placido system has inherent
limitations as the systems must make shape assumptions
that while reasonable in normal corneas are inaccurate
in abnormal or pathologic corneas where there are non-
linear changes in curvature.[4] Many devices also contain
qualitative classification systems and quantitative indices
and algorithms for data interpretation. A sample printout
of one of the Placido-based devices giving an elevation map
on the lower left depicted in Figure 7.
However, CVK does have limitations: There is a lack of
standardization between instruments; it depends on reference
axis [Figure 8], alignment, and focus; it is susceptible to artefact
(distortion, tear film effect); it is based on simplified optics
(only applies to central cornea); and there is a smoothing
effect as explained below. Also sampling occurs around
the circumference of the mires, there is no measurement
between mires.
It is important to understand the first limitation more
carefully as it will affect all the measurements on Placido-based
devices. The figure [Figure 9] explains the relations of the
various axes with respect to the apex of cornea and its effect
on the image acquisition. According to the American National
Standards Institute (ANSI) standard definition, the corneal apex
is the point of maximum curvature on the cornea, whereas
the vertex is the point nearest to the camera of the Placido
instrument located on the corneal topographer axis (CT axis).
Before acquisition, the topographer aligns this axis normal
to the cornea. Pseudokeratoconus patterns can be created
when line of sight (the line passing through the fixation
Figure 7:AtypicalprintoutofthePlacido-baseddeviceshowingaxial
(upperleft), tangential/ instantaneous(upperright), elevation (lower
left),andwavefrontmaps(lowerright)
Figure 8:Curvaturetopographyisreferenceaxisdependent.Thegure
ontheleftshowsanonaxistopographyshowingasymmetricbowtie
pattern.Thismeasurementwhenrepeatedwithslightdecentrationwith
respecttothelineof sight(blackline)showsanasymmetricbowtie.
CourtesyProfBelin
object nodal point and the fovea), the corneal apex, and the
videokeratoscope normal (CT axis) do not line up.
Most of us visualize the eye as a Gullstrand-reduced eye,
assuming that the eye is symmetric, with the line of sight
coinciding with the visual axis, and crossing the center of pupil
and corneal apex. This, however, is not always the case.[5,6] More
so, we assume that the measurement axis of the Placido system
also coincides with the above. Most people do not look through
the center of their cornea.[7,8] A person with pseudo-strabismus
due to large angle kappa (angle between pupillary axis and the
visual axis) demonstrates these principles. The person looks as
though their eyes are not straight (their line of sight does not go
through the corneal apex which is the point on the cornea having
the maximum curvature), but when you perform a cross-cover
test, the eyes are straight (there is no re-fixation movement).
When you perform a Hirschberg test, however, the reflected
light appears displaced. This is because a reflected image (same
as in a Placido videokeratoscope) needs to align normal to the
corneal surface to appear straight. When the apex and the line
of sight differ, the reflected image appears abnormal (in the
adult imaged on a Placido videokeratoscope, this would appear
as an asymmetric bowtie Figure 8], but the eye is still physically
normal. This is the problem with trying to reconstruct shape from
a curvature measurement. There are other methods of depicting
curvature (i. e., instantaneous or local) that obviate some, but not
all, of the above limitations. Sagittal (axial) curvature, however,
remains the most commonly used. Figure 10 shows how the
effect of this decentration is less in the elevation-based devices.
The standard topographic curvature (axial or sagittal curvature)
is a referenced-based measurement. It is not a unique property of
the cornea. The same shape can have many different “curvatures”
depending on which axis is used to make the measurement.
“Keratoconus” maps can be created on Placido devices
in normal aspherical surfaces with angular decentrations
as small as 5 degrees. This again underlines the need to
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1 49
have more information about the corneal surface through
multiple technologies and also without forgetting basic
clinical examination like the Hirschberg tests and cover tests.
The important differentiating feature among the normal
and the abnormal corneas in this scenario would be the
orthogonality of astigmatism, normal pachymetry, stable
refractions, and best corrected acuity of 20/20 in spite of
having an asymmetric bowtie pattern.[9] Because clinicians are
less familiar with interpreting curvature data, these devices
convert this information to power values with the paraxial
formula (P = (n-1)/r; where P = corneal power, n = 1.3375
(compensates for negative power of posterior cornea by
incorporating a fixed correction in the refractive index to
compensate for posterior corneal power), and r = radius of
curvature in meters). This ignores spherical aberration but is
a good approximation for the power of the central cornea.
Curvature maps are usually displayed in one of two
formats — axial or tangential — depending upon what
method is used to calculate the radius of curvature. For
the axial curvature map, r = the distance from the corneal
surface to the optical axis along the normal (vertex normal)
and all radii are measured from this axis. Due to this common
reference axis, small irregularities may not be visible or
“smoothened out” as they are very small as compared to the
large corneal diameter. The axial maps represent a running
average of scaled curvatures. Extreme values are averaged out
of the calculation. These maps are spherically biased and are
calculated on the assumption that all rays of light striking the
corneal surface are refracted, forcing a focal point through the
optical axis as a reference axis [Figure 11a]. This is similar to a
keratometer and assumes that the center of rotation of the best
fit sphere lies on the optical axis. It is a good approximation
for the paracentral cornea (2-mm zone).
The axial map is the most commonly used and provides a
good estimate of overall corneal shape, which appears smooth
with little noise because it provides an average of adjacent
curvature values. This is useful for evaluating corneal optics
(i. e., central power of cornea, calculating intraocular lens (IOL)
power, and screening for pathology). Axial map, also referred
to as sagittal maps and can be converted to “Refractive” maps
applying the refractive index, Snells law, and ray tracing
instead of Gaussian optics that is used for the axial map. This
map plots the refractive power of the cornea at each point.[10]
This accounts for spherical aberration outside the central
zone, and provides information about the imaging power of
the cornea. This is helpful for correlating curvature to vision
and analyzing surgical effects.
On the other hand, for the tangential (local, instantaneous)
map, r = the instantaneous radius of curvature at each point
on the cornea. This is the true “r,” independent of the defined
central axis, and is therefore a more accurate measure of
curvature at a particular point. As a result, the tangential
(instantaneous) map is noisy because it is more sensitive to
local changes and accentuates focal abnormalities. This is
useful for evaluating corneal shape (i.e., ectasia, assessment
of refractive surgery candidates, surgically induced changes,
and contact lens fitting). The difference in the two types of
maps is illustrated with example in Figure 11b.
Figure 12 is an elevation map for one of the Placido-based
device. These maps are however may not be very accurate and
reliable as the assumptions (sphero-cylindrical optics) used in
their construction do not perform well in the setting of non-
linearly altered corneal shape (ectasia/post-surgery).[4] These
devices derive the elevation maps using the angle of reflection,
Figure 9:Ingure,thelineofsightisdifferentfromthecornealapexA,
thiswouldresultinanartifactonPlacido-basedimagingandexplains
thedisplacedapexsyndrome
Figure 10:Thesagittalcurvaturemap (upper right) shows a gross
asymmetric bowtie pattern (which is considered abnormal) but the
frontand thebackelevation mapsshownbelow arenormal. This is
characteristicsofdisplacedapexsyndromeinwhichthelineofsight
andthecornealapexdonotcoincide leadingto asymmetricbow tie
patterninotherwisenormalcorneas.CourtesyProfBelin
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
50 Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1
whereas true elevation can only be measured accurately by
triangulation method employed in slit scan and Scheimpflug
devices.
In addition to the type of map display, the map scale
(dioptric range, step size, number of colors) is also very
important because it affects sensitivity. An absolute scale is
constant for all exams and is useful for comparisons over time
and between patients. A relative or normalized scale adapts
to the range of powers on the corneal surface and differs
for each cornea. Thus, the power range and step size may
be narrow or broad, which magnifies or minifies significant
changes. Sensitivity is also affected by the step size (dioptric
range for each map color). The recommended step size is 1.5 D
(Universal standard scale).[11] Small steps increase sensitivity
by adding more colors and exaggerate minor or normal
changes, which can cause confusion (i. e., pseudo keratoconus)
and misdiagnosis. Large steps decrease sensitivity and mask
significant changes due to smoothing of points between rings.
Topographic artefact can occur with inappropriate step size,
misalignment with the CT axis, pressure on the globe, and
altered tear film.
Most CVK instruments also contain quantitative measures,
indices, and algorithms to aid in data evaluation. The most
commonly used machine classifiers (artificial intelligence) are
the KISA devised by Rabinowitz and Rehman, the keratoconous
prediction index (KPI) by Maeda et al., Cone location and
magnitude index (CLMI) among the many others that are
commercially available. The classical Rabinowitz system and
the Klyce and Maeda systems are available on most commercial
Placido devices (e. g, TMS). They have different methods of
assessment, in that the Rabinowitz system has sharp defined
cut-offs and is made with data from keratoconus patients to
differentiate it from normals and relies on examination of
both eyes. It does not account for other causes of topographic
abnormalities.[12] KISA is a composite index calculated as
KISA% = (K) × (I-S) × (AST) × skewed radial axis index
(SRAX) × 100. In this case, only absolute values are used
without sign, any K value less than 47.2 was substituted by
1 and only ones in excess of 47.2, difference used (if K is 57.2,
the value input is 10). SRAX is the difference between 180 and
smaller of the two angles between the radii and the AST index
quantifies the degree of regular corneal astigmatism (Sim K1-
Sim K2). Using this index set at 100 percent in eyes with no
other pathology in their study, Rabinowitz et al., detected no
overlap between normal eyes and keratoconus.[13] According
to the authors, this score also has a range for suspects with
minimum overlap to the normal population. Sixty to one
hundred percent values are considered to be suspect. The KPI
index developed by Maeda and Klyce consider eight different
indices together to give a result and help differentiate
keratoconus from other corneal irregularities.[14] The CLMI
is a novel index in that it is platform independent and can
be used with any machine. It can track the location and
magnitude of cone over time to assess progression. It showed
100% specificity and 100% sensitivity in separation of cones
and normal eyes when used on a validation set.[15] All these
software however are no substitute for history and thorough
clinical examination and also an independent validation in
new set of population.
These softwares calculate the probability of the patients
map resembling a keratoconic or other known patterns
of anterior curvature by comparing it using a pre-defined
mathematical model fed into the machine. Figure 13 shows a
model flow chart for the data acquisitions to calculations of a
Figure 11:(a)Thisgureshowsthedifferenceintheconcept ofthe
sagittalandthetangentialcurvature.Asseenintheguretheradiiof
curvatureintheaxialmaponleftarederivedfromthedistancesfrom
thereference(opticalaxis)andthederivationofthetangentialcurvature
isindependentofthereferenceaxisasshowningureonright(b)This
showsactualmaps toillustratetheprinciplesexplained in11a. The
mapsbelongtothesameeyeofthesamepatientbutdifferinvalues
andappearanceasthederivationisinadifferentmanner.Tangential
mapontherightmakeslocalizedelevationlookmoreobvious
a
b
Figure 12:Figureshowstheelevationmapfortheanteriorsurfacein
Placido-baseddevicetheleftuppercornerindicatedthereferencebody
anditsdimensionsthatareusedforthecurrentcalculations
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1 51
Placido-based device for sake of explanation. The performances
of these classifiers in clinical practice are tested and evaluated
depending on their receiver operating characteristics curves
involving their parameters and indices. The specificity and the
sensitivity of these classifiers in detecting true keratoconus
should be known for each one individually, before one starts
to use these in clinical practice and would be well-advised
to examine these before hand in case a device purchase is
contemplated, as they vary with each machine.
However, the importance of looking at the raw data, its
quality and the topographic map can never be undermined
and is most important before the interpretation of any report
is started. Machine classifiers use mathematical models
developed from the regression analysis of the patients with
keratoconus and other irregularities and try to fit the data
acquired for every patient into them, to give a spontaneous
response in relation to the acquired data. Caution must
be exercised in the stand-alone use of these indices for
interpretation of topography printouts.
Relevance of Elevation-Based Devices:
(Tomographers)
The occurrence of iatrogenic keratectasia after uneventful
LASIK on normal eyes with no traditional risk factors (age,
pre-op corneal thickness, ablated depth, residual stromal bed,
normal Placido reports, and single point corneal thickness
values) has demonstrated the need to know more about the
cornea undergoing refractive surgery.[16] In addition, several
recent studies have demonstrated the role of the epithelium in
masking the cone on Placido-based devices.[17-19] The accuracy of
multiple standard point ultrasound pachymetry in predicting
change in corneal shape is found wanting as pointed out by
Rabinowitz et al.,[20] Epithelial changes around the cone tend to
mask the early ectatic changes on the anterior surface. Inability
of the Placido topographers to give accurate anterior elevation
maps in all scenarios as cited before definitely necessitates
a need to search for new and accurate tools to obtain more
information on the cornea. Modifications in the curvature,
asymmetry, and elevation differences in the posterior surface
have been well-documented in keratoconus eyes[21-23] These
studies using different instruments reported greater posterior
astigmatism, posterior elevation, and prolacity in suspect eyes
as compared to normal. The consensus around the discriminant
values is however lacking, highlighting the fact that the
patterns are to be more relied upon than the actual “values”
derived. Hence, the current weight of evidence suggests that
the study of the posterior surface appears to be invaluable to
the decision-making in refractive surgery. These facts without
doubt stress the vital role of elevation-based devices in corneal
evaluation pre-refractive surgery.
Elevation-Based Devices (Tomographers)
Elevation-based topographers are the relatively new
entrants into the market and use different principles for
Figure 13:ThisgureillustratestheowchartfortheprototypePlacido-baseddevicefromtheimageacquisitiontotheneuralnetworkandthe
naloutcome
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
52 Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1
acquisition and calculation vis a vis the Placido-based
devices. These devices are basically of three types: The
Slit scanning devices, the Scheimpflug devices, and the
OCT-based devices. In these devices, the machine usually
provides a composite printout display. The most common
maps that are provided in their display include the anterior
elevation map, the posterior elevation map, sagittal power
display, and the pachymetric map. The posterior elevation
and the pachymetry map are the valuable additions as
compared to the Placido-based devices.
The understanding of the elevation-based devices should
start with the concepts of elevation and reference surface.
Concept of Elevation
In terrain topography, the surface elevation is surveyed in
reference to sea level which is fixed. The localized elevations
in the cornea being small relative to the cornea itself, to
unmask these irregularities the global curvature must
first be eliminated akin to the pattern standard deviation
calculations in visual fields. This can be achieved by fitting
the cornea with a surface with features most resembling it,
the so-called reference surface. This reference surface can be
a sphere, asphere, ellipsoid, toric aspheroid, etc. Although the
cornea is not exactly spherical in shape, the most commonly
used reference surface used is a sphere. Not many studies in
literature have addressed the best choice of reference shape
to detect ectatic change but it should be remembered that
this choice can affect the final image/output significantly.
In one study, the use of the toric ellipsoid was associated
with decreased risk of masking the cone as compared to
the best fit sphere (BFS).[24] Another group using the Galilei
analyzer has recommended the use of the toric aspheroid
had the ability to differentiate in between the Forme fruste
keratoconus (FFKC) and the normal corneas using the Galilei
analyzer.[25] The rationale behind this finding is that by virtue
of being very close to the natural corneal shape, it will
highlight the abnormalities better which otherwise would
be hidden in the ridge pattern normally generated by the
BFS. Familiarity with the different outputs obtained with
the respective reference surfaces in varying scenarios would
help us identify suspicious patterns in each of them over a
period of time. Individual effort on the part of each refractive
surgeon is desired to use each of the reference surfaces in
the best possible way.
Elevation and its Relation to Slope
and Curvature
These are three different measurements of a surface. Corneal
elevation is the measurement of height between points
at two different elevations and is primary source data for
tomographers. For the Placido-based devices, it is the “Slope”
which measures steepness or incline between two points
that is derived by the analysis of reflections of rings on the
cornea. Slope is the first derivative of elevation and can be
used to either calculate elevation or curvature. Curvature is
calculated from slope data in Placido topography system and
elevation is calculated from it using integral mathematics.
The Scheimpflug prototype (tomographers) Pentacam uses
elevation as primary data and calculates curvature by
differential mathematics. Since surgically altered surfaces
can have non-linear changes in curvature, a method that
“calculates” and not actually measures elevation may be
inaccurate due to multiple elevation (mathematical) solutions
available for these surfaces.[26]
Float
The reference surfaces can be fitted in the two ways as depicted
in Figure 14. The float is a method of unconstrained fit where
no pre-condition are defined for the position/locations of the
reference surface. Here, the reference surface fits the surface to
be measured with minimum square difference that is minimum
difference above as well as below the measured surface. In the
pinned or apex fit, the center of the reference is pinned on the
view axis and the apex of the reference surface is located on/
pinned to the surface to be studied as shown in the Figure 14b.
This is in short a “constrained” fit, where conditions are
imposed on the reference surface. As a result the pinned fit
being located with its apex on/pinned to the surface to be
studied and not elevated about it lowers the central hill (for
prolate surfaces ) seen on the float fitting due to its different
methodology. The float method is the commoner of the two
and what is normally shown by all elevation instruments
nowadays. This method basically fits the reference surface to
the surface in question with minimum square difference. The
fitting type should be kept in mind while looking at different
instruments as it makes a difference in the output as shown
in Figure 15. One should bear in mind that the “float” is not
synonymous with “elevation” or elevation map and is actually
a method of fitting of a reference surface in elevation-based
devices.
Figure 14:Types oftting methodsfor areferencesurface.1Apex
t/center+pinned—Centerofreferenceobjectisconstrainedonthe
viewaxisanditintersectsdatasurfaceontheviewaxis.Thisattens
thecentralhill as it centers on it,Float— Center is unconstrained.
Referencetsthe corneal surface withminimumsquare difference.
Almostalldevicesusethismethodasithastheleasterror
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1 53
Best Fit (BF) Surface
The actual raw data obtained by the elevation-based
topographers lacks qualitative patterns that would allow the
clinician to easily separate normal from abnormal corneas.
In other words, raw elevation data for normal eyes look very
similar to the raw elevation data in abnormal eyes as shown
in the Figure 16.
To give a qualitative definition to the elevation data the
machine using the above concepts of elevation and float,
identifies the dimensions of a selected reference shape that
can best fit to the examined surface for each eye tested
depending on its individual characteristics. This calculated
reference shape varies in dimensions for each eye and its shape
and curvatures are indicated on the printout. This is called as
the best fit reference surface. This surface is fitted in a pre-
defined “Fit zone” which is usually 8 mm in diameter is used
in most machines. Most of the software in different machines
have their specific reference surface setting (e. g., the Belin
Ambrosio display (BAD) has the BFS). In other situations, one
can use different references depending on individual choice
and experience.
Scanning Slit System
Orbscan is the prototype machine of this type.
This was the first attempt at the study of the posterior
surface of the cornea which started in 1995 with the
introduction of Orbscan I. Later the Placido was added in the
second version of the device. The scanning slit system was
introduced for the first time with the Orbscan that used the
theorem of slit scan triangulation for the calculation of the
corneal power. Triangulation is a highly accurate mathematical
concept in use in satellite navigation and land topography to
calculate distances using fixed known reference and is used in
this device. The diagram [Figure 17] highlights the principle
of scanning slit imaging and triangulation.
The device uses the projection of a slit of light [Figure 18a] at
various positions on the vertical meridian on the cornea, takes
images at pre-specified positions using the video camera, and
calculates the curvature at these positions using triangulation.
The entire cornea is covered with 40 vertical slits, 20 on each
side normal to the surface at each position of acquisition,
capturing the backscattered light with the video camera. Each
of the slits has 240 data points.[27] The data for the area between
these slits is interpolated. The approximate acquisition time is
Figure 15:The gure illustrates theeffectof the ttingmethodson
thenalresult.Thegureontheleftutilizedtheoatandtheoneon
therightdidnot
Figure 17: This gure shows the principle of scanning slit imaging
andtriangulation
Figure 16:Figureshowsthattherawelevationdatafrom normalas
wellasabnormaleyeslacksqualityandlookssameifnotcompared
toreferencebody(CourtesyProfBelin)
Figure 18: Orbscan which uses the projection of a slit of light, (a)
showstheprojectedslitoflightonthecorneaandbshowstheOrbscan
machineinsideviewandfrontview
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
54 Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1
1.5 seconds. In addition, the device has a Placido disc for the
calculation of the anterior curvature Figure 18b. This Placido
disc however is not circular and cannot get data from the upper
and lower parts of the cornea consistently.
Many ophthalmologists and researchers have proposed
their own scales and indices for interpretation of the Orbscan
images. The readers are advised to refer to specific literature on
the device for the same. Figure 19 shows a prototype Orbscan
quad map image that is most commonly used and the basic
parts of the printout are explained here.
In many of the instrument reviews and research, certain
problems with the Orbscan have been highlighted vis a vis the
Scheimpflug devices. In normal corneas, the pachymetry lags
in accuracy and reproducibility to the Scheimpflug devices[28]
and the readings were not interchangeable.[29,30] The Orbscan
inaccurately identifies the post-operative posterior corneal
surface and routinely locates the surface too anteriorly. As a
result, its pachymetry reading is too thin, and its topography
suggests ectasia. The inability of the Orbscan to identify the
posterior corneal surface on post-LASIK eyes promoted the
false beliefs that changes to that surface were commonplace
after LASIK and that most patients exhibited subclinical
ectasia.[6,27,31-33]
Although pre-operative pachymetry is repeatable and
correlates well with ultrasound after a built-in “fudge factor,”
the Orbscan IIz’s measurements of the central corneal
thickness after myopic LASIK are less than those measured
by ultrasonic pachymetry. This difference decreases with time
and may not be significant after 1 year. Figure 20 shows some
common drawbacks on the Orbscan printout.
Edge-detection algorithms that are the heart of the scanning
slit-based Orbscan IIz system are vulnerable to interference
from artifacts introduced by the corneal reshaping. The
Orbscan IIz in particular exaggerates the posterior corneal
surface’s contour, and clinicians must be careful to avoid an
over-interpretation of this topographic analysis for several
months following refractive surgery.[34]
Due to the above reasons, the Scheimpflug technology
appears to supercede scanning slit devices in their accuracy
and usage in refractive surgery clinics nowadays.
Scheimpug Imaging
The concept of Scheimpflug photography was started by
an Austrian naval forces officer, who was a cartographer by
profession, Theodore Scheimpflug. This was initially used for
the purpose of topographic imaging for military purposes
with cameras attached to the gliders or hot air balloons with
the prototype devices. His method of photography could
correct for the perspective distortion of the aerially acquired
photographs [Figure 21].
The initial concept used multiple fixed cameras that used
to capture images. This was replaced by the moving charge-
coupled device (CCD) camera in the ophthalmic devices to serve
the same purpose as illustrated in Figure 21b. The images so
acquired are subjected to analysis and reconstruction to give
the information for the surface studied. Figure 22 illustrates
the principle of Scheimpflug imaging vis a vis the conventional
camera.
Aptly called as corneal “tomography,” this imaging modality
involves the acquisition of slices of the cornea (each one
passing diametrically through the center) by a rotating camera
and their analysis. The technology of these devices is totally
different from the Placido-based devices and they derive the
elevation data as their primary raw data. This data can be
Figure 19: shows the various indices that need to be seen while
inspectinganOrbscanmapforevidenceofkeratoconus.Allfourmaps
1.Anteriorelevation2.Posteriorelevation3.Keratometricmap(Power
map)4.Thicknessmap(pachymetry)needtobeseenindetailalong
with the indices 5. anterior and posterior best t sphere 6. various
Indices(simKs,keratometricasymmetryin3-and5-mmzone,pupil
diameter,anglekappa,etc.)
Figure 20:Showsthe drawback of the keratometric map lower left
inthatthe map is truncatedaboveand below due to thehorizontal
shapeofthemires(asshowninthephotoofthedeviceinFigure18).
The pachymetric readings of the upper and lower parts have not
been acquired and there is no quality score for the image in this
particular device. One should refrain from interpreting images with
incompletedata
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1 55
converted by means of advanced mathematical algorithms into
curvature data. As opposed to attempting to generate elevation
data from curvature (integral), the calculation of curvature
from elevation data provides a unique solution (differential).
Of the Scheimpflug devices that are available in India, the
Pentacam is the only “purely Scheimpflug” device and the
other two, the Sirius and the Galilei are bimodal devices that
combine the large cone Placido with the Scheimpflug. The
Galilei has 2 Scheimpflug cameras for faster image acquisition
and image averaging, whereas the Pentacam and the Sirius
have 1 each [Figure 23]. The commonalities of the devices are
that they acquire the cross-sections of the surface illuminated
by slit beams in various meridians with the help of rotating
Scheimpflug cameras and analyse them. As all the illuminated
slits projected pass through the center of the cornea, the
central measurements are accurate.
One pertinent difference between the Pentacam and
the other two devices is that these two derive the anterior
curvature by a combination of Placido and the Scheimpflug
data by the use of proprietary software and by incorporating
the refractive index value that is mentioned on the printouts.
The Pentacam derives the curvature purely form the elevation
data using mathematical methods and readings may not
strictly be always interchangeable between these devices for
the sheer difference in the technology they use. This has been
documented by various comparative studies to that effect. The
curvature data derived by the Pentacam has compared well
with the Placido in the studies done till date in normal as well
as abnormal eyes.[35,36] The Galilei and the Pentacam HR (The
version of Pentacam with high-resolution Scheimpflug camera
and uses 1,20,000 points) are devices capable of analyzing
more than 1,00,000 data points from high definition images.
The other available devices use fewer data points.
The most important advantage of the tomographic devices
(Scheimpflug, OCT, and slit scan imaging) is the global
pachymetric map and the posterior elevation map. [Figure 24]
This data is possible only with the above technologies (and
the OCT that is not discussed here). The Scheimpflug scores
over the slit scanning devices due to better edge identification
and reproducibility of data as has been proved by various
publications.[27,28,3133] Commonly, the clinician views elevation
data not in its raw form (actual elevation data) but compared
to some reference shape. Figure 25 shows the Scheimpflug
raw data (images).
Figure 26 shows the concept of the reference surface (in
bold) located under the surface to be evaluated. In these
elevation maps, it is important to remember that the colors
represent the elevation data. Any point on the cornea that is
higher than the best-fit reference surface will be shown as a
peak — in the “hotter” colors, and any point that is lower than
Figure 23: These are the various Scheimpug devices commonly
available.1)PentacamwithsingleScheimpugcamera2)Siriuswith
Placido(large)and the Single camera (yellow) 3) The Galilei-Dual
cameradevicewithalargeconePlacido
Figure 22:ThisisanillustrationtoshowhowtheScheimpugcamera
principleworks(right)withrespect totheconventionalcamera(left).
Thismethodofimageacquisitionenhancesthedepthoffocus
Figure 24:Globalpachymetrymap(left)andthebackelevationmap-
(right)importantadvantagesoftheelevation-baseddevicesoverthe
Placido-basedsystems
Figure 21: (a) Photo graph of the original Scheimpflug camera
(b)180/360-degreerotationofthecameraaroundthecorneaimaging
theperpendicularilluminatedcornealslits
a b
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
56 Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1
the best-fit sphere will be shown as a valley — in the “cooler”
colors. [Figure 26]. In the printouts, the reference surface, its
diameter/s, method of fitting used, and the fitted area are
mentioned in addition to the color scales used as shown in
Figure 27. The elevated points are given (+) values and the
depressed points are given (−) values as shown in Figure 28.
These reference shapes as discussed before can be spheres,
aspheres, toric aspheroids, and ellipsoids with multiple options
available and customisable in every device. The maps typically
display how actual corneal elevation data compares to or
deviates from this known shape. The map is obtained for the
anterior as well as the posterior elevations. The output data
will depend on the choice of the reference surface [Figure 29]
and one may well follow the manufacturers’ guideline and
personal experience to set the reference surface and interpret
the results accordingly.
Fitting a reference surface to the central 8.0-mm zone
appears best, as most of the pathology lie inside this zone,
this provides adequate data points and most users should be
able to obtain maps without extrapolated data. Most of the
instruments have their acquisition set within the 9-mm zone.
Since the normal eye is a surface aspherical prolate fitting the
central 8-mm zone allows for subtle identification of both
ectatic disorders and astigmatism. Use of larger zones leads
to more artefacts due to the aspheric prolate nature of the
corneal surface.
Ectasia and keratoconus are diseases that involve
thinning of the cornea and hence the pachymetric maps are
invaluable evidence to that effect. The value of pachymetric
progression has been demonstrated by Luz et al., where it
is seen that the pachymetric variation from limbus to the
thinnest point in normal eyes is distinct from eyes with
keratoconus.[37] Indices of curvature and thickness that are
generated centered around the thinnest point can detect
mild forms of keratoconus undetected by Placido-based
neural network program.[38] The Belin Ambrosio (Enhanced
ectasia) display [Figure 30] in the Pentacam incorporates
these novel parameters as percentage thickness increase (PTI)
Figure 25:Rawdatasetofthepentacam
Figure 27:Thegureshowsthenotationforreferenceobjectsbeing
usedforthecalculation.Theellipsoidreference(BFTEF)isusedinthis
caseandhasamaxandmincurvaturedened.Thettingmethodis
oatandtheareattedis8mm
Figure 26:Thisgureillustratesthe colorcodinginelevation-based
topographyinasimpliedway. Thelevelofthereferencesurfaceis
shownasayellowline.Theareaelevatedabovethereferencesurface
isshowninredandtheareadepressedbelowisshowninblue
Figure 28:Theplussignwithhotcolorsandtheminussignwithcool
colorsindicatesthelocationofthepointsaboveandbelowthereference
surface,respectively
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1 57
from thinnest point and the Corneal thickness spatial profile
(CTSP) This map has a normal and a hyperopic database in its
latest version. Another feature of this software is the ability
to enhance the cone location. This is done by subtracting
the 4mm area around the thinnest point and calculating the
new BFS for the rest of the cornea (which would be flatter
if the cone is located in the excluded area). As a result when
the excluded area is compared with the flatter “new” BFS, it
stands out if abnormal in the “enhanced map” that is shown
at the bottom of the printout for both the anterior and the
posterior surface. In addition to these features, the display
in its current third version (BAD3) incorporates the K max,
maximum front, and back elevation in microns, a pachymetry
map, thin point location, displacement of the thin point from
apex, and a pachymetry-based classifier the ART max. Besides
this machine classifier, the main classifier, the “D” value,
incorporates 9 parameters for its calculation and has been
independently validated in a retest population.[39]
The curvature maps are the indirect/surrogate indicators
to thinning and protrusion and may not match the accuracy
of the elevation devices in predicting the change of corneal
shape as they do not image the posterior cornea and do not
provide a thickness map. According to some researchers also,
the anterior curvature changes are initially masked by the
thinning/heaping of the epithelium forming a typical donut
pattern (Central thinning over the cone with peripheral ring
of thick epithelium) and hence the posterior elevation and
thickness profile assumes peculiar importance.[17]
The sagittal or axial curvature maps are poor indicators
of the location of the cone in keratoconus and commonly
exaggerate its peripheral appearance. Both anterior elevation
maps, posterior elevation maps, and pachymetric maps more
accurately locate the true cone position.
The various devices incorporate various type of machine
classifiers that can give a mathematical representation of
the acquired data and classify it into normal and abnormal
patterns as is [Figures 31a and b]. The classifiers would be
as good as the data that is presented to them for calculation.
Hence, the quality and the technique of acquisition of images
would be of paramount importance for a good result.
Interpretation of Topography/Tomography
Reports
Regular astigmatism shows a classic pattern where the flat
meridian is depressed with respect to reference surface and
the steep meridian is above or elevated with respect to the
reference surface in tomography maps and shows a symmetric
bowtie with orthogonal axes on topography. The larger the
astigmatism the greater the difference between corresponding
points on the principal meridians. Additionally, the further you
go out from the center or apex the greater the deviation from
the reference surface. Irregular astigmatism is by definition
where the principal meridians are non-orthogonal. This
is readily apparent in the maps. Mild changes may still be
associated with good best spectacle corrected vision (BSCVA),
but larger amounts of irregular astigmatism are typically
associated with a reduction in BSCVA.
Irregularly, irregular corneas are so distorted that the
principal meridians can often not be identified. These corneas
are almost always pathologic, associated with a significant
reduction in BSCVA and may be seen in conditions such
as advanced keratoconus, Pellucid degeneration, anterior
dystrophies, and corneal scarring.
In screening for refractive surgery cases, in addition to a keen
sense of detail in topographic patterns, it is important to have
sensitive and specific indices to minimize the false positives as
well as false negatives for detecting problematic corneas. Each
technology and software has advantages as well as limitations.
Different technologies alone or in combination that study
the various aspects of the cornea to the fullest will offer the
greatest sense of security for selecting a proper case. Given
the constraints of all currently available devices and classifiers
examination of a doubtful case with more than one system would
be prudent choice. A complete overview of the technology and the
mathematics used is paramount to select a good method/s for our
Figure 29:Thisshowsbackelevationofthesameastigmaticeyewith
differentreferencesurfaces. The toric ellipsoid to left of thesphere
totheright.Thepatternofastigmatismisbestseenonthespherical
referencesurface.Thescaleontheleftshowsthecolorcodingusedfor
theabovemaps.Asseenherethetoricsurfacemoresnugglytsthe
surfaceinquestionleadingtosmallervaluesofelevation/depression
Figure 30:Belin Ambrosioenhancedectasia display(BAD)version
3.E.g.,Keratoconusmap
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
58 Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1
practice. Population validations of machines and their different
statistical tools that are published in literature add authenticity
to them and inspire confidence in their usage.
Before looking at the interpretation of the maps we have
to understand what we are looking for.
Keratoconus and FFKC
Keratoconus is a clinical diagnosis and FFKC is a subtle
topographic abnormality before clinical manifestation of the
disease.
The aim of topography and tomography in refractive
surgery clinic is to rule out keratoectatic disease either
in form of frank keratoconus or subtle FFKC as they are
contraindications to the procedure. There are certain topo
and tomographic criteria for both obtained through the work
of various researchers and it would be worth a mention just
before proceeding to interpret the reports. The suspicious signs
for keratoconus include:
Axial map abnormalities
1 K greater than 48 D.
2. SRAX greater than 21 degrees.
3. I-S greater than 1.42D.
4. Corneal astigmatism on anterior odr posterior surface
greater than 6 D.
5. Against the rule astigmatism.
6. S-I difference at the 5-mm zone >2.5 D.
On the elevation maps
1. Isolated island or tongue-like extension on either surface
(BFS mode).
2. Elevation values greater than 12 microns on the anterior
elevation map in the central 5 mm (BFTE mode).
3. Elevation values greater than 15 microns on the posterior
elevation map (BFTE mode).
Pachymetry/corneal thickness map: On Scheimpug devices
1. Thinnest location less than 470 microns.
2. Displacement of the thinnest point >500 microns from the
center.
3. Pachymetry difference asymmetry in two eyes at thinnest
point >30 microns.
4. S-I difference at the 5 mm circle >30 microns.
5. Cone-like pattern on the thickness map.
Steps in Interpretation of the Topographic
Maps for Refractive Surgery Screening
The ideal protocol will depend on the devices and their
availability, whether a Placido or a Scheimpflug scanning
slit or both are available. Since the global pachymetry and
the elevation of the posterior surface are available only on
the Scheimpflug, OCT or scanning slit they would provide
more information and would logically even be superior if the
curvature maps obtained from them are comparable to the
Placido-based devices. Hence, the elevation-based devices
especially the Scheimpflug-based ones would be a better
choice, after all the publications and the physics discussed
above. Besides, there are more false positives and negatives
in Placido-derived images also. The devices like the Pentacam
have demonstrated to have not only comparable but also
interchangeable[36] results (with Placido) for the anterior surface
and can be used as standalone device and so can be the bimodal
devices (Sirius, Galilei) that contain both technologies.
Interpretation of topography printouts is NOT all about
pattern identification but also looking in between the lines.
An important error to be avoided in all instances is to jump to
the machine classifier results directly and basing your clinical
decision on them. The step wise interpretation of the reports
would include:
Follow the sequence of GRADES:
G- General information.
R- Reliability (Quality).
A- Abnormal/Normal.
D- Defect.
E- Evaluate.
S- Subsequent test.
1. Patient demography, eye, date of procedure, and which eye
is being examined.
2. Look at the basic data on the map, note the quality of raw
data (mires in Placido and the edges in Scheimpflug), so also
Figure 31:Theguresaandbshowstheexamplesofvariousmachine
classiers for the different devices and different types of printouts.
Thereisalargevarietyofprintoutspossibledependingonthesoftware
installedinthemachine
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1 59
the quality scores that are available in the devices. On the
display, correlate this to the acquisition area and watch for
areas of missing data (data gaps), indicated by hatchings
or absent data. If poor quality is seen (in terms of image
appearance or a poor quality score), repeat the images or
select a better image. In case one does not procure the
images himself/herself, it is prudent to see the Placido raw
image selected for calculations especially if abnormal. The
edges of the Scheimpflug map should be seen for hatchings
that come within the central 8-mm zone/areas of absent
data (as the case may be depending on the instrument
make).
3. It is pertinent to start by looking at the reference scales that
are in use on your machine unless the steps are kept constant
(using absolute scales/universal scale, etc.) as a rule. If the
current data has to be compared to a previous report done
elsewhere, similar scales can be used to get an approximation
preferably on a device with similar technology. Steps of 1.5
D are the usual standard for curvature data/2.5 µm for
pachy/10 µm for eleviation (contour). Absolute scales allow
comparison between successive examinations of same
patients and over different patient groups.
While using the Placido-based devices stand alone, the basic
patterns for the anterior curvature on the axial map need to
be compared to the standard patterns provided by Rabinowitz
et al.,[12] and Levy et al.,[40] [Figure 32]. The map patterns can be
classified into circular, oval, steepening (superior or inferior),
bowtie (symmetric and asymmetric), and with or without
skewing of the radial axes, J and the inverted J as shown in the
Figure 32 template. The symmetrical bowtie, round, and the
oval are considered normal, the asymmetric bowtie, skewed
axes, inferior steepening, and J and inverted J pattern, and
their various permutations as suspicious. The Pellucid (crab
claw), butterfly, and the keratoconus (D) patterns are examples
of abnormal patterns.
Look for the maximum keratometry of the anterior surface
(Kmax) and correlate with the ultrasound corneal thickness
values (measured at multiple points).
Use the instantaneous map to look for the location of
maximum power/cone location. The corneal thickness above
this location can be mapped with ultrasound and compared
to a symmetric location on the other side of the pupil. The
difference more than 30 microns is suspicious.[10] This is
followed by examining the machine classifiers like the KISA,
KPI, etc. available with the Placido-based devices. If available,
the wavefront maps and the modulation transfer function can
be correlated at this stage along with the history of patients
refractive changes over the years.
4. Look at the elevation maps (for elevation-based Scheimpflug
and scanning slit devices) — If a Scheimpflug or scanning
slit is being used, the elevation data and the global
pachymetry can be obtained from the Scheimpflug or
the slit scan device itself. The important points to be
noted in the elevation maps are the posterior elevation
patterns, thinnest point location, the peripheral corneal
thickness values (and the percentage thickness increase
(PTI) if available as in Pentacam) in addition to the anterior
curvature data (acquired independently in bimodal devices
through Placido or derived mathematically as in Pentacam).
5. The most elevated points on the anterior and the posterior
elevation maps should be correlated to the highest power
on Axial/Saggital curvature map and the thinnest point
on the global pachymetry map. If all the above match,
it is called as the “fourpoint touch” and is a hallmark of
suspect cornea, especially if the apex is decentered by more
than 500 microns and the peripheral thickness readings
of the upper and lower half at the 7-mm zone also show
a significant difference of greater than 100 microns. Look
over each index and the values provided by the device like
the K max and other specific indices.
6. These indices should be used with clinical correlation to the
patients’ demographics (i. e., Younger patient with suspect
topo/tomography more significant as compared to an older
individual with same changes), inter eye asymmetry (in
patterns, axes, elevations, pachymetric behavior, maximal
corneal power), and the refractive error and put into
perspective for consideration in refractive surgery.
7. After examining all the above, the statistical statement
(or machine classifier report eg, KISA, KPI, BAD “D value,”
etc.) provided by the respective machines can be seen and
correlated to the above findings. Some of these indices have
undergone independent validation in population-based
studies and are more robust than others. The maps will also
provide a final comment on its analyses of the presented
data and flag it (as normal or abnormal or as percentage
probability of abnormalcy) depending on the softwares that
are built into them.
8. If it is a repeat test, compare it with the previous tests and
note the changes, especially if a suspect cornea is being
followed up. Most of the machines have an in-built program
Figure 32:ThesearethebasicshapetemplatesgivenbyRabinowitz
andLevytodepictthechangesintheaxialmapsinvariouscorneas.
Levystudiedthefamiliesofpatientsofkeratoconustoseethepattern
inrstdegreerelatives.(Jandinversejpatterns)asymmetricbowtie
withskewedradialaxesandinferiorsteepening;andJinv,asymmetric
bowtiewithskewedradialaxesandsuperiorsteepening
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
60 Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1
to generate comparative difference map to highlight the
changes.
It is important to remember that these machine classifiers have
their own limitations in terms of specificity and sensitivity and
are not to be considered as the gospel truth. A thorough search
of validation of these classifiers by the individual physician
by looking up peer reviewed journals is recommended before
putting them into practice.
Decision-making with the topography/tomography reports
Putting corneal evaluation into refractive surgery practice.
Having interpreted the reports carefully and seen the
various indices with the available technology to evaluate
the topographic risk, the further assessment would
include patient factors like desired correction, age, inter
eye asymmetry, calculated residual stromal bed following
ablation, and the pre-operative corneal thickness. One such
system to predict the risk of an individual eye and allot a
risk score was proposed and validated by Randlemann in
2008.[41] It contained the use of five parameters (Topography,
residual stromal bed, Age, Corneal thickness pre-op, and
Manifest refraction spherical equivalent (MRSE) and led
to a cumulative score that would be predictive of the risk.
This system however had its own share of criticism and
shortcomings about the sensitivity,[42,43] its accuracy in the
setting of normal topography result[44] and scoring design.
But till such a time that a new system becomes available by
continued research and incorporation of other risk factors
Figure 33:Keratoconusmapexemplifyingthe4pointtouch.Inthisgure,thethinnestpointofthecorneacoincideswiththehighestpoweron
thetangentialmapandthemostelevatedpointontheanteriorandposteriorsurfaces.Thisisclassicalofectaticdisease
Figure 34:Thisgureshowsafrontsurfacecurvaturemapacornealthicknessmapandfrontelevationmap.Sagittalmapshowsnearlysymmetric
bowtiewithminimalskewingoftheaxiswhichisalsoreectedontheelevationmap.Thisischaracteristicsoftheastigmatism
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1 61
like tomographic indices it may used as a reasonable guide
for decision making.
In one of the recent published literature, the percentage
of the (preoperative) corneal thickness ablated (including
thickness of the flap) was the risk factor most predictive
of ectasia risk among all others.[45] Thus, the knowledge of
topography and its correct application forms the vital cog in
the wheel of successful refractive surgery practice.
Important Glossary
1. Curvature axial: This is the commonest map used in
topography it is a running average of the corneal
power and used more commonly in IOL power
calculations.
2. Curvature tangential (instantaneous): This map measures
local irregularities better and is commonly used in
refractive surgery to detect suspect corneas.
3. Refractive power map: IT is a map depicting the various
points in diopteric power.
4. Elevation Map (anterior and posterior elevation) maps
that show the deviation of the examined surface from the
utilised reference surface.
5. Raw elevation data: It is the data which is not usually
displayed in elevation base devices but used for
calculations due to its lack of qualitative nature.
6. Best fit surface: It is that surface that is used for generating
elevation maps and can be manually or automatically
fitted to the surface in question using different algorithms
like float or apex fit.
7. Best fit sphere: It is a spherical reference surface that
best fits the measured surface by the different fitting
algorithms.
8. Float: It is an algorithm to fit the reference surface to the
surface in question using minimum square difference.
9. Apex fit: Is the constrained fitting of the reference surface
with the center on the view axis and intersecting the
examined surface on the axis.
10. Wave front map: It is Zernike or a fourier analysis of
the examined surface and is available in most of the
topographer devices and helps in understanding higher
order aberrations.
Above in Figures 32-36 e. g. are given for exercise.
References
1. Pan CW, Cheng CY, Sabanayagam C, Chew M, Lam J, Ang M,
et al. Ethnic variation in central corneal power and steep cornea
in Asians. Ophthalmic Epidemiol 2014;21:99-105.
2. Chua J, Tham YC, Liao J, Zheng Y, Aung T, Wong TY, et al.
Ethnic dierences of intraocular pressure and central corneal
thickness: The Singapore epidemilology of eye diseases study.
Ophthalmology 2014;121:2013-22.
3. Markaikis G, Roberts CJ, Harris JW, Lembach RG. Comparison
of topographic technologies in anterior surface mapping of
keratoconus using two display algorithms and six corneal
topography devices. Int J Kerat Ect Cor Dis 2012;1:153-7.
4. Schultze RL. Accuracy of corneal elevation with four corneal
topography systems. J Refract Surg 1998;14:100-4.
5. Ara RC, Klyce SD, Busin M. Keratometry in refractive surgery.
J Refract Surg 1986;2:6.
6. Rubin ML. Optics for Clinicians. Gainesville: Triad Publishing
Company; 1993.
7. Mandell RB. Evere Kinsey Lecture. The enigma of the corneal
contour. CLAO J 1992;18:267-73.
8. Ara RC, Warnicki JW, Rehkopf PG. Corneal topography using
rasterstereography. Refract Corneal Surg 1989;5:414-7.
9. Belin IC 92 handout. ESCRS. Vienna; 2011.
10. Holladay JT. Keratoconus detection using corneal topography.
J Refract Surg 2009;25:S958-62.
11. Smolek MK, Klyce SD, Hovis JK. The Universal Standard Scale:
Proposed improvements to the American national standards
institute (ANSI) scale for corneal topography. Ophthalmology
2002;109:361-9.
12. Rabinowitz YS. Keratoconus. Surv Ophthalmol 1998;42:297-319.
13. Rabinowitz Y, Rasheed K. KISA% index: A quantitative
videokeratography algorithm embodying minimal topographic criteria
for diagnosing keratoconus. J Cataract Refract Surg 1999;25:1327-35.
14. Maeda N, Klyce SD, Smolek MK. Comparison of methods for
detecting keratoconus using videokeratography. Arch Ophthalmol
1995;113:870-4.
Figure 35:Asymmetricbowtiewithskewingofaxesandsteepening.
Thismapshowscentralelevatedareawithasymmetricbowtiepattern
andskewingoftheradialaxisAB(SRAX)
Figure 36:Nearovalpatternwhichisanormal
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
Dharwadkar and Nayak: Topography and tomography
62 Journal of Clinical Ophthalmology and Research - Jan-Apr 2015 - Volume 3 - Issue 1
15. Mahmoud AM, Roberts CJ, Lembach RG, Twa MD, Herderick EE,
McMahon TT. CLEK Study Group. CLMI: The cone location and
magnitude index. Cornea 2008;27:480-7.
16. Ambrósio R Jr, Dawson DG, Salomão M, Guerra FP, Caiado AL,
Belin MW. Corneal ectasia aer LASIK despite low pre operative
risk: Tomographic and biomechanical findings in the unoperated,
stable fellow eye. J Refract Surg 2010;26:906-11.
17. Reinstein DZ, Archer TJ, Gobbe M. Corneal epithelial thickness
profile in the diagnosis of keratoconus. J Refract Surg 2009;25:
604-10.
18. Gatinel D, Racine L, Hoang-Xuan T. Contribution of the corneal
epithelium to amterior corneal topography in patients having
myopic photorefractive keratectomy. J Cataract Refract Surg
2007;33:1860-5.
19. Touboul D, Trichet E, Binder PS, Praud D, Seguy C, Colin J.
Comparison of front surface corneal topography and Bowman
membrane peculiar topography in keratoconus. J Cataract Refract
Surg 2012;38:1043-9.
20. Rabinowitz YS, Rasheed K, Yang H, Elashoff J. Accuracy of
ultrasonic pachymetry and videokeratography in detecting
keratoconus. J Cataract Refract Surg 1998;24:196-201.
21. Schlegel Z, Hoang-Xuan T, Gatinel D. Comparison of and the
correlation between anterior and the posterior corneal elevation
maps in normal eyes and keratoconus suspect eyes. J Cataract
Refract Surg 2008;34:789-95.
22. Tomidokoro A, Oshika T, Amano S, Higaki S, Maeda N, Miyata K.
Changes in the anterior and posterior corneal curvatures in
keratoconus. Ophthalmology 2000;107:1328-32.
23. Piñero DP, Alió JL, Alesón A, Escaf Vergara M, Miranda M. Cornea
volume, pachymetry and correlation of anterior and posterior
corneal shape in subclinical and dierent stages of keratoconus.
J Cataract Refract Surg 2010;36:814-25.
24. Kovács I, Miháltz K, Ecsedy M, Németh J, Nagy ZZ. The role
of reference body selection in calculating the posterior corneal
elevation and prediction of keratoconus using the rotating
Scheimpug camera. Acta Ophthalmol 2011;89:e251-6.
25. Smadja D, Santhiago MR, Mello GR, Krueger RR, Colin J,
Touboul D. Inuence of the reference shape for discriminating
between normal corneas, subclinical keratoconus and keratoconus.
J Refract Surg 2013;29:274-81.
26. Belin MW. AAO 2006 monograph. The Pentacam: Precision,
confidence, results and Accurate Ks.
27. Cairns G, McGhee CN. Orbscan computerized topography:
Aributes, applications, and limitations. J Cataract Refract Surg
2005;31:205-20.
28. Lackner B, Schmidinger G, Pieh S, Funovics MA, Skorpik C.
Repeatability and reproducibility of central corneal thickness
measurement with Pentacam, Orbscan and ultrasound. Optom
Vis Sci 2005;82:892-9.
29. Guilbert E, Saad A, Grise-Dulac A, Gatinel D. Corneal thickness,
curvature and elevation readings in normal corneas: Combined
Placido Scheimpug v/s Placido scanning slit system. J Cataract
Refract Surg 2012;38:1198-206.
30. Crawford AZ, Patel DV, McGhee CN. Comparison and repeatability
of the keratometric and corneal power measurements obtained
by Orbscan II, Pentacam, Galilei corneal topographers. Am J
Ophthalmol 2013;156:53-60.
31. Charkrabarti HS, Craig JP, Brahma A, Malik TY, McGhee CN.
Comparison of corneal thickness measurements using ultrasound
and Orbscan slit-scanning topography in normal and post-LASIK
eyes. J Cataract Refract Surg 2001;27:1823-8.
32. Giessler S, Duncker GI. Orbscan pachymetry aer LASIK is not
reliable. J Refract Surg 2001;17:385-7.
33. Cairns G, Ormonde SE, Gray T, Hadden OB, Morris T, Ring P,
et al. Assessing the accuracy of Orbscan II post-LASIK: Apparent
keratectasia is paradoxically associated with anterior chamber
depth reduction in successful procedures. Clin Experiment
Ophthalmol 2005;33:147-52.
34. Cairns G, McGhee CN, Collins MJ, Owens H, Gamble GD. Accuracy
of Orbscan II slit-scanning elevation topography. J Cataract Refract
Surg 2002;28:2181-7.
35. Mauger TF, Cynthia JRoberts, Lena V Chheda, Rebecca A
Kuennen, Hendershot AJ. Comparison of placido scheimpug
and combined dual scheimpflug- placido technologies in
Evaluating anterior and posterior CLMI, Sim K s as well as K max,
keratoconic and post refractive surgery Ectasia. Int Keratoectat
Corneal Dis 2012;1:44-52.
36. Khachikian S. Consistent curvature: Assessing the cornea March
2009. Eyeworld newsleer.
37. Luz A, Ursulio M, Castañeda D, Ambrósio R Jr. Corneal thickness
progression from the thinnest point to the limbus: Study based on
normal and keratoconus population to create reference values. Arq
Bras Ohalmol 2006;69:579-83.
38. Saad A, Gatinel D. Tomographic and topographic properties of
forme fruste keratoconus corneas. Invest Ophthalmol Vis Sci
2010;51:5546-55.
39. Villavicencio O, Fatimah Gilani, Maria A Henriquez, Luis Izquierdo
Jr, Renato R Ambrosio Jr, Michael W Belin. Independent population
validation of the Belin Ambrosio enhanced ectasia display:
Implications for keratoconus studies and screening. Int J Kerat
Ect Cor Dis 2014;3:1-8.
40. Levy D, Hutchings H, Rouland JF, Guell J, Burillon C, Arné JL,
et al. Videokeratographic anomalies in familial keratoconus.
Ophthalmology 2004;111:867-74.
41. Randlemann JB, Traler WB, Stulting RD. Validation of the Ectasia
Risk Score System for preoperative laser in situ keratomileusis
screening. Am J Ophthalmol 2008;145:813-8.
42. Chan CC, Hodge C, Suon G. External analysis of the Randleman
Ectasia Risk Score System: A review of 36 cases of post LASIK
ectasia. Clin Experiment Ophthalmol 2010;38:335-40.
43. Duey RJ, Hardten DR, Lindstrom RL, Probst LE, Schanzlin DJ,
Tate GW, et al. Ectasia aer Refractive Surgery 2008;115: 1849.
44. Binder PS, Traler WB. Evaluation of a risk factor scoring system
for corneal ectasia aer LASIK in eyes with normal topography.
J Refract Surg 2010;26:241-50.
45. Santhiago MR, Smadja D, Gomes BF, Mello GR, Monteiro ML,
Wilson SE, et al. Association between the percent of the tissue
altered and post laser in situ keratomileusis ectasia in eyes with
normal pre operative topography. Am J Ophthalmol 2014;158:
87-95.
Cite this article as: Dharwadkar S, Nayak BK. Corneal topography and
tomography. J Clin Ophthalmol Res 2015;3:45-62.
Source of Support: Nil. Conict of Interest: None declared.
[Downloaded free from http://www.jcor.in on Wednesday, September 28, 2016, IP: 91.15.87.159]
