SMART PHONE BASED FUNDUS CAMERA

1. DESIGN OF OPTICAL SYSTEM FOR SMART PHONE
BASED FUNDUS CAMERA
A PROJECT REPORT
Submitted by
HARISH S
RAM PRATAP V
SUBAMURUGAN V
AJAY RATHINAM
in partial fulfilment for the award of the degree
of
BACHELOR OF ENGINEERING
in
MECHATRONICS ENGINEERING
COIMBATORE INSTITUTE OF ENGINEERING AND TECHNOLOGY
(AUTONOMOUS)
ANNA UNIVERSITY: CHENNAI 600 025
APRIL 2021

2. ANNA UNIVERSITY: CHENNAI 600 025
BONAFIDE CERTIFICATE
Certified that this project report “DESIGN OF OPTICAL SYSTEM FOR
SMART PHONE BASED FUNDUS CAMERA” is the bonafide work of
HARISH S 710518115012
RAM PRATAP V 710518115026
SUBAMURUGAN V 710518115030
AJAY RATHINAM R 710518115301
who carried out the project work under my supervision.
Signature
Dr. K. MANIKANDA
SUBRAMANIAN, M.E., PhD
HEAD OF TH DEPARTMENT
Professor
Department of Mechatronics
Engineering
Coimbatore Institute of Engineering
and Technology, Coimbatore 641
109
Signature
K. SENTHIL KUMAR, M.E.
SUPERVISOR
Assistant Professor
Department of Mechatronics
Engineering
Coimbatore Institute of Engineering
and Technology, Coimbatore 641
109
Submitted for the University Project Viva-Voce held on _________________
Internal Examiner External Examiner
I

3. ACKNOWLEDGEMENT
First of all, we extend our heartfelt gratitude to Dr. K.A. CHINNARAJU
Director and the Management of Coimbatore Institute of Engineering and Technology
for providing us with all sorts of supports in the completion of this project.
We record our intentness to Dr. N. NAGARAJAN M.E., Ph.D., Principal, Coimbatore
Institute of Engineering and Technology for his guidance and encouragement for the
successful completion of this Project.
We are highly grateful to Dr. K. MANIKANDA SUBRAMANIAN M.E.,
Ph.D., Professor and Head, Department of Mechatronics Engineering, for his valuable
Suggestions and guidance throughout the course of this project. His positive approach
had offered incessant help in all possible ways from the beginnings.
We are highly grateful to our project guide Mr. K. SENTHIL KUMAR M.E.,
Assistant Professor-I, Department of Mechatronics Engineering, for his Valuable
Suggestions and guidance throughout the course of this project. His positive approach
had offered incessant help in all possible ways from the beginnings.
We also extend my sincere thanks to all the faculty members of Department of
Mechatronics Engineering, Parents and friends who have rendered their Valuable help
in completing this Project successful.
II

4. ABSTRACT
The ever-increasing popularity and availability of smartphones
and the rapid advances in technology for capturing and sharing
images with them have resulted in the expanding use of
smartphones as a clinical imaging device in ophthalmology. This
application has been facilitated by the ease of use and portability
of the smartphones and the already
extensive mobile-phone networks, and it presents a unique
opportunity for applications such as telemedicine and self-
diagnosis. Retinal photography (fundus photography) is an
essential part of ophthalmology practice. Acquisition of high-
quality fundus images requires a combination of appropriate
optics and illumination usually in the form of a condensing lens
and a coaxial light source. This is the reason that a commercial
fundus camera costs tens to hundreds of thousand dollars. We
describe in detail a relatively simple technique of fundus
photography in human eyes using a smartphone, an app for
android mobiles, and instruments that are readily available in an
ophthalmic practice.
III

5. TABLE OF CONTENTS
CHAPTER NO. TITLE PAGE NO.
BONAFIDE
ACKNOWLEDGEMENT
ABSTRACT
1 INTRODUCTION 1
1.1 Project overview 1
1.2 Project architecture 3
1.3 Indirect ophthalmoscopy 3
1.4 Using of indirect ophthalmoscopy 3
1.5 Working with patient 4
1.6 Performing the exam 5
1.7 Mastering the use of IDO 5
1.8 Fundus image and refractive error 7
2 LITERATURE SURVEY 9
3 PROJECT OBJECTIVE 11
4 HARDWARE 12
4.1 Optical lens system 12
4.1.2 Customized lens specification 12
4.2 Illumination system 13
IV

6. 4.2.1 Quantitative Analysis of wide field Fundus
image 14
4.2.2 Methods trans-pars-planar illumination 14
4.2.3 Experimental Setup 16
4.3 Design of bellow arrangement 20
4.4 Imaging sensor 20
5 SOFTWARE DESCRIPTION 22
5.1 Zemax optic studio 22
5.2 Image simulation output 24
6 PROJECT DESCRIPTION 26
6.1 Working principle 26
6.1.1 Image Capturing 26
6.1.2 Image Processing Phase 26
6.1.3 Diagnosis 27
6.2 Project design 27
6.3 Procedure 27
6.4 Calculation and measurement 28
6.4.1 Dimension of Eye 29
6.5 Output images 29
6.6 Experimental setup 30
7 CONCLUSIONS 31
8 REFERENCES 32
V

7. 1
CHAPTER 1
INTRODUCTION
1.1 PROJECT OVERVIEW
In India most of the people are suffering from different eye diseases like
cataract eye, glaucoma, diabetic retinopathy, night blindness etc. According to the
survey of eye diseases in urban and rural populations around 62.9% of different ages
are suffering with eye diseases.
In this paper we discussed a smart phone-based screening technique called
Fundus Photography and Retinal Ophthalmoscopy. Fundus cameras are a key part
of any ophthalmic practice’s equipment, as they can be used to locate and document
a variety of ocular defects with relative ease. As common as they are though, fundus
cameras are oftentimes bulky and expensive pieces of equipment that may not be
feasible for use in smaller offices, preventing many populations who may be at risk
from getting the help they need. Additionally, most systems are mydriatic systems
which means that they require the patient’s eye to be dilated, placing further
restrictions on the likelihood of smaller practices investing in a system. The creation
of a low-cost, portable, non-mydriatic system would allow for the device to become
much more accessible, allowing for more practices and organizations to own a
device and perform tests to document and diagnose more patients. This is especially
important with the worldwide prevalence of diabetic retinopathy, which can occur
as a result of uncontrolled diabetes and may lead to blindness if left untreated. The
implementation of a low-cost, portable system would allow for at-risk populations
to get examined for the defect and determine if they need to seek treatment, helping
to reduce the risk of blindness. In 1926, Carl Zeiss Company introduced the first
commercially available fundus camera, which offered a 10.8 retinal field and
required manual exposure using flash powder and colour film.1,2 Since then, the
capabilities of fundus cameras have improved dramatically to include non-mydriatic
imaging, electronic illumination control, automated eye alignment, and high-
resolution digital image capture. These improvements have helped make modern
fundus photography a standard ophthalmic practice for detecting and documenting
retinal disease. Although current fundus cameras have advanced significantly since

8. 2
their introduction, the traditional tabletop optical design has remained largely
unchanged. Complex optical assemblies in current devices provide high-resolution
imaging of the fundus but also require dedicated clinical space and high
manufacturing costs. Portable cameras have recently become commercially
available, but most remain difficult to use in a hand-held manner and often have
substandard image quality, compared to their tabletop counterparts. The commercial
field of fundus camera equipment stands in unique contrast to consumer digital
camera technology, where personal cameras are becoming ever cheaper, smaller,
and easier to use. Although other ophthalmic equipment manufacturers have recently
incorporated consumer digital single-lens reflex (DSLR) cameras into their fundus
camera designs, they do not make full use of the consumer camera’s built-in
functions or space-saving design. Traditional fundus camera designs are thus ill
suited to leverage the significant cost reductions and technological advancements of
consumer camera technology.
Within the past decade, retinal screening programs for common eye disease,
such as diabetic retinopathy and age-related macular degeneration, have experienced
rapid growth. The expansion of these screening programs into rural, nurse operated,
highly distributed primary care facilities highlight the importance of having access
to an inexpensive, portable, easy-to-operate, and high-image-quality fundus camera.
Our goal was to create a device capable of imaging the human fundus and
documenting retinal pathology with components that cost less than Rs 75,000. We
also aimed to improve dramatically the ease of use of the device by incorporating
common ‘‘point-and-shoot’’ consumer camera technology. A secondary objective
was to reduce the design to a portable form factor that would enable remote use of
the device in settings such as hospital bed consultations and nursing home facilities.
This design would provide a means of acquiring fundus photographs in clinical
settings previously inaccessible to tabletop cameras.

9. 3
1.2 PROJECT ARCHITECTURE:
Hardware
Software
1.3 INDIRECT OPHTHALMOSCOPY:
The technique of examining the fundus of the eye is called ophthalmoscopy.
In direct ophthalmoscopy, a virtual and erect image of the fundus is seen. In
indirect ophthalmoscopy, a real and inverted image is formed between the
condensing lens and the observer. The advantage of stereopsis (depth perception)
and a larger field of view makes indirect ophthalmoscope (IDO) more useful both
in retina clinics and during posterior segment surgeries. In this project our fundus
camera is designed based on indirect ophthalmoscopy.
1.4 USING OF INDIRECT OPHTHALMOSCOPY
1. Alignment: Put the indirect on, and ensure your oculars and light spot are
properly centered. While focusing the light spot at your hand at arm length,
close one eye at a time to make sure your pupillary distance is properly
adjusted and you can see well with each eye.

10. 4
2. Adjust the brightness: Don’t go crazy on the brightness (60-80% is generally
enough on most models). If too bright, you will often spend more time
fighting a patient’s Bell’s reflex (tendency for eyes to roll upwards when
trying to shut them).
3. Choose your spot size: If the patient’s pupil is wide and dilated, use the
largest spot size. If the pupil is mid dilated, use the medium spot size. If the
pupil is small, use the small pupil size. The reason for this is that when light
hits the iris and reflects back, it creates glare and makes it harder to discern
retinal structures.
4. Your hand positioning: Hold the lens with your thumb and index finger and
plant a pinky finger on the patient’s forehead or cheek. If you’re using bigger
lenses (e.g., 20D), you will have to hold the lens further away from the
patient’s face than smaller lenses (e.g., 28D).
5. Your head positioning: All beginners reflexively move their head closer to
the lens and the patient in order to try to see better. Fight this urge! Many of
our learners get an “aha” moment when they realize you need to be
appropriately far enough from the lens in order to get the light to focus
properly.
1.5 WORKING WITH THE PATIENT
● Angling the patient: It’s hard to examine a patient while they are sitting up.
Initially, try to lay the patient back at 45-60 degrees to make the distances
easier to manage and try to stand directly opposite of where you are looking.
E.g., If you are looking at the left temporal macula, stand on the patient’s
right.
● Turn their head: Have the patient turn their head slightly towards you,
whichever side you are standing on.
● Dealing with noses: When a patient’s nose gets in the way (like when you are
standing on the patient's left side examining their right temporal macula) have
them turn their head more towards you. They can still move their eyes in
whatever direction you need them to, but it moves the nose out of the way.
● Giving instructions: When describing directions for patients to look, it is
sometimes easier to tap on their face than to give a direction (down and
right), and makes your exam more efficient.

11. 5
1.6 PERFORMING THE EXAM
1. Starting the exam: Start with a peripheral view (have the patient look up) as
this will help acclimate a patient to the light. If you start the exam by looking
at the macula, your patients will be angry.
2. Visualizing the periphery: To look in the far periphery, tilt your own head 45
degrees to the left or right. At that angle the pupil effectively becomes
elliptical and you can fit the indirect illumination beam and one ocular into
that ellipse.
3. Beginner tip: Use the diffuser light on the indirect to help illuminate the far
periphery when you’re first starting out. It makes the alignment of the lens
less critical.
1.7 Mastering the use of IDO
The technique of indirect binocular ophthalmoscopy starts with good
dilatation of pupils with tropicamide and phenylephrine drops and examining fundus
in the darkroom. Before fundus examination, adjust your interpupillary distance and
make sure the light spot is well centered from both oculars at arms distance. To
visualize the posterior pole, the examinee will be asked to look directly into the light
source. Peripheral quadrants are visualized by asking the examinee to look in the
direction of the quadrant to be examined and the observer standing diagonally
opposite to the quadrant of interest. For example, to visualize the superior fundus of
the eye, the examinee is asked to look superiorly (towards the examinee's forehead)
and the observer stands towards the foot end of the examinee.
Fig 1.1 Eye Screening

12. 6
The fundus is examined systematically (superior [SUP], super-temporal quadrant
[STQ] temporal [TEMP], inferotemporal quadrant [ITQ], inferior [INF],
inferonasal quadrant [INQ], nasal [NAS], super nasal quadrant [SNQ], and lastly
posterior pole) and one eye after the other. (Figure-6b). To note, the aerial image
observed is the image of the retina from the quadrant of interest but it is reversed
and inverted. Tilting the lens to avoid unwanted reflexes and moving the
condensing lens towards the examinee’s eye or observer’s eye allows full
illumination of the lens with the aerial image.
Fig 1.2 Position of Observer and Fundus Examiner
The peripheral most of the retina and pars plana region bows inward making direct
visualization of this portion difficult. Depressors are needed to indent sclera pushing
the peripheral portion inward to bring an area of interest into focus. This manoeuvre
is often needed in poorly dilating pupils or while examining retinal breaks in the
retinal periphery. Good dilatation obviates the need for scleral indentation and
associated discomfort. Depression should be gentle and smooth over the eyelid
avoiding the tarsal plate of both lids. The examinee should rotate his/her eye towards
the quadrant to be examined and the tip of the depressor should indent the area of
interest. (Figure:7) Importantly make sure the depressor, the examinee gaze, and
observer must be along the same axis to visualize the indented portion. When the

13. 7
fundus image is difficult to visualize, the novice should check this alignment as an
initial step.
Fig 1.3 Scleral indentation technique
1.8 FUNDUS IMAGES AND REFRACTIVE ERRORS
Knowledge about image formation in the different refractive status of the eye
is important for effective visualization. Magnification and extent of an aerial image
depend on the refractive status of the eye. The fundus image is larger in hyperopic
at the cost of a lesser field of view; this is more in case of an aphakic eye. In contrast,
Myopic fundus details appear smaller with a larger field of view. Images in
emmetropic eyes with posterior intraocular lens appear nearly 2% larger than their
phakic counterpart and corresponding slightly reduced field of view. In the oil and
gas-filled eyes, the overall refractive status changes and so, magnification and field
of view changes accordingly.
Fig 1.4 Gas and Oil filled eyes
In gas-filled phakic eyes, the posterior surface of the lens acts as a high concave lens
causing a myopic shift. In Aphakia, the posterior surface of cornea acts as a high

14. 8
convex lens neutralizing anterior concave surface, thus visualizing fundus without
condensing lens. Oil in phakic makes the posterior surface of lens low minus causing
hyperopic shift and convex oil bubbles in aphakic causes myopic shift bringing down
aphakic hyperopia. (Figure 8) Depending on the distance at which an aerial image is
formed, the aerial image is brought into focus by moving the condensing lens closer
or away from the examinee's eye.
Ophthalmoscopy in paediatric eyes needs special mention. Procedures need
to be gentle and quick before the babies/children turn uncooperative. Smaller babies
are swaddled to immobilize and small size speculum is used to keep lids separated.
Indenters are used to rotate the eyeball for a full glance of the fundus. Low
concentration dilating drops (tropicamide 0.5% - 1% with phenylephrine 2.5%) are
preferred and not to forget blocking the punctum while instilling dilating drops.

15. 9
CHAPTER 2
LITERATURE REVIEW
KENNETH TRAN, THOMAS A. MENDEL, is from an international
journal discussing the topic of “Construction of an Inexpensive, Hand-Held Fundus
Camera through Modification of a Consumer Point-and Shoot Camera”. This paper
provides an overview of construction of non-mydriatic fundus cameras. The detailed
study of construction of a low-cost prototype model of fundus camera is obtained
from this paper. It also provides comparative study assessing the image quality of
the prototype camera against a traditional tabletop fundus camera conducted under
an Institutional Review Board (IRB)-approved study.
G. LI, H. ZWICK, B. STUCK D.J. LUND from Walter Reed Army
Institute of Research is described about “On the use of schematic eye models to
estimate retinal image quality”. The paper discussed different schematic eye models
depending on the accuracy of measurement of the indices of refraction, curvature,
and thickness of ocular components. From this paper we design the schematic human
eye model for lens design in Zemax. It also provides details like modulation transfer
function and cut-off frequency.
MAHMUT KARAKAYA from “16th
annual conference of midsouth
computational Biology & Bioinformatics Society”, is discussed about the topic of
“Comparison of smartphone-based retinal imaging systems for diabetic retinopathy
detection using deep learning” in this paper, smart phone based retinal imaging
systems available in market is discussed. We can find comparative results of
different manufacturers like iNview, D-Eye, iExaminer, and peek-retina etc.
We can observe that most companies use iPhone instead of android mobiles
which results in expensive. So, we use android mobile for cost efficiency.
BENQUAN WANG, XINCHENG YAO, R.V. PAUL CHAN from
research paper in scientific reports is discussed about “Contact free trans-pars-
planar illumination enables snapshot fundus camera for non-mydriatic wide field
photography” in this paper complete information about trans-pars-planar

16. 10
illumination used in conventional fundus photography. This illumination system
illuminates light through the pars plana, an area outside of the pupil. It also describes
how to create a prototype instrument that can achieve 90 degree fundus view
coverage in single shot fundus images, without the need of pharmacological pupil
dilation was demonstrated.
DEVRIM TOSLAK, YANJUN CHEN, MUHAMMET KAZIM EROL
from “international research journal published in HHS Public Access” is
discussed about “wide-field fundus imaging with trans-palpebral illumination” in
this paper describe about palpebral illumination system in which 90 dioptre
ophthalmic lens and a 25 dioptre relay lens are used. In this adequate illumination
level was obtained to capture wide angle fundus images within ocular safety limits,
defined by ISO standards. This novel trans-palpebral illumination approach enables
wide angle fundus photography without eyeball contact and pupil dilation.

17. 11
CHAPTER 3
PROJECT OBJECTIVE
The main objective of this project is to mass eye screening of groups of people
in rural areas. In this, we use Fundus Camera to capture retinal images and run it
through software for image processing. The processed image is used for examining
the patient, whether he/she is having any retinal diseases. This data is stored in
software applications.
1. To design an optical lens system for capturing retinal images.
2. To set up an illumination system that supports the optical system.
3. To manufacture a below arrangement that consists of both optical and
illumination system which is attached to an android mobile device.
4. To design a model which is advanced to the previous model but also should
be cost effective to support non-mydriatic systems.
5. The device helps in examination of common eye diseases – Cataract,
Glaucoma, Diabetic retinopathy, Age related macular degeneration, & Dry
eye.

18. 12
CHAPTER 4
HARDWARE DESCRIPTION
4.1 OPTICAL LENS SYSTEM
We design a sequential mode of optical system which is used to capture
retinal and cornea images. This system consists of different types of lenses which
are arranged in a sequential manner. Mostly we used plano-convex and bi-convex
lens in this system. The system is designed and simulated in Zemax optic studio
software. The lens is manufactured in a customised design. We design this lens
system based on optical calculations to obtain 50 degrees of field of view.
4.1.1 Customized Lens Specification
Fig 4.1 Lens editor data
We manufactured and arranged the optical setup of the lens system according to
the above-mentioned specifications. We use AR glass coating to avoid geometrical
aberrations. We mostly prefer Aspheric lenses to avoid geometric aberrations

19. 13
caused by the curved surface of the object plane. The whole lens is arranged in the
below setup which is integrated with the android mobile.
4.2 ILLUMINATION SYSTEM
Contact-free trans-pars-planar illumination enables a snapshot fundus camera
for non-mydriatic wide field photography. In conventional fundus photography,
trans-pupillary illumination delivers illuminating light to the interior of the eye
through the peripheral area of the pupil, and only the central part of the pupil can be
used for collecting imaging light. Therefore, the field of view of conventional fundus
cameras is limited, and pupil dilation is required for evaluating the retinal periphery
which is frequently affected by diabetic retinopathy (DR), retinopathy of prematurity
(ROP), and other chorioretinal conditions. We report here a non-mydriatic wide field
fundus camera employing trans-pars-planar illumination which delivers illuminating
light through the pars planar, an area outside of the pupil. Trans-pars-planar
illumination frees the entire pupil for imaging purpose only, and thus wide field
fundus photography can be readily achieved with less pupil dilation. For proof-of-
concept testing, using all of-the-shelf components a prototype instrument that can
achieve 90° fundus view coverage in single-shot fundus images, without the need of
pharmacologic pupil dilation was demonstrated.
Wide field fundus photography is desirable for screening, diagnosis, and treatment
evaluation of diabetic retinopathy (DR)1,2, retinopathy of prematurity (ROP)3,4 and
other eye diseases that can produce morphological abnormalities at peripheral areas
of the retina. Traditional fundus cameras employ trans-pupillary illumination, i.e., a
donut-shaped illumination pattern projected to the peripheral area of the pupil. After
passing through the pupil, the light diverges and illuminates the posterior of the eye5.
To illuminate the retina homogenously, the diameter and divergence of the
illumination pattern on the pupil plane must be carefully adjusted, requiring careful
design and sophisticated construction of the optical imaging system6,7. According
to ISO 10940:20098, external-angle is commonly used to specify field of view
(FOV) in traditional fundus cameras. However, eye-angle has been recently adopted
to determine the FOV in wide-field fundus imagers, such as a Retinal cam (Natus
Medical Inc., Pleasanton, CA), Optos (Optos Inc., Marlborough, MA), etc. In order
to avoid unnecessary confusion, we provide both external- and eye-angle numbers
in the following discussion. Traditional fundus cameras provide 30°–45° external-
angle (45°–67.5° eye-angle) FOV9. Additional challenges with trans-pupillary

20. 14
illumination include glare caused by light refection from the cornea and crystalline
lens5,7 , and the requirement of adequate pupil dilation for wide field examination.
Pharmacologic pupil dilation may make patients suffer from light glare and focusing
difficulty for hours and even days in some cases.
4.2.1 Quantitative analysis of wide field fundus image
Quantitative analysis of fundus images is essential for objective and
automated classification of eye diseases14. In order to verify the potential feasibility
of using the trans-pars-planar illumination-based fundus camera for quantitative
imaging, we explored automated classification of arteries and veins, quantitative
analysis of blood vessel diameter and tortuosity, and arteriolar-to-venular diameter
ratio (AVR). It is known that retinopathy can affect arteries and veins differently.
For example, some studies have shown that in ROP the increase in arterial tortuosity
is more significant than that of veins15, and in DR the diameter of arteries decrease
and diameter of veins increase16,17. Therefore, separate analysis of arteries and
veins can provide improved sensitivity for quantitative fundus image analysis and
classification. Figure 2 illustrates basic procedures of automated classification of
retinal arteries and veins. Technical details are explained in Methods section. First,
red and green channels were separated from a colour fundus image (Fig. 1b2).
Second, the green channel was used to segment individual blood vessels in Fig. 2a
to reconstruct the blood vessel map (Fig. 2b). Third, the optical density ratio (ODR)
between red and green channels was calculated18. As shown in Fig. 2c, arteries
showed lower ODR value than veins. Fourth, a brightness threshold was applied in
Fig. 2c to separate arteries and veins (Fig. 2d). the automated classification
reasonably matches manual classification of arteries and veins. Figure 2e shows
average diameters of arteries and veins. The AVR thus could be calculated as
AVR=194 μm/235 μm = 0.8, which is within normal range (0.54–0.82) reported in
previous publication19. Figure 2f shows calculated artery and vein tortuosity.
4.2.2 Methods Trans-pars-planar illumination
Figure 4 shows the anatomy and location of the pars planar. The pars-planar
is the smooth, posterior part of the ciliary body (Fig. 4a). It is a ~4mm wide band
located ~3–4mm posterior to limbus22, 23. the pars planar lacks muscle, blood
vessels and pigmentation23, thus is more transparent than other area of the sclera,
making it a good location for delivering light into the eye. Figure 4b shows an

21. 15
infrared image of the eyeball. the eyeball was illuminated with infrared light
(850nm) at the other side using a fibre that touched eyelid. As shown in Fig. 4b, the
brightness of the pars planar was higher than other clear areas because of its high
transmittance. Figure 4. Schematic (a) and photographic (b) illustrations of Pars-
planar. (a) Schematic cartoon showing the cross-section of the human eye. Pars-
planar is the area labelled between the dashed red lines. (b) Near infrared photograph
showing the location of pars plan, marked by dashed red lines. Figure 5 shows
schematic illustration of trans-pars-planar illumination in comparison with other
illumination schemes. In conventional fundus cameras, illumination and imaging
light paths are spatially separated to avoid refection from cornea and crystalline lens,
i.e., the peripheral area of the pupil is used for illumination and the central area is
used for imaging, to achieve reflection-free fundus imaging. As shown in Fig. 5a1,
a donut shaped illumination pattern is projected to the pupil plane. the illumination
light diverges after passing the pupil and illuminates the posterior area in the eye
(Fig. 5b1). As shown in Fig. 5b1, such configuration requires a large pupil so that
illumination and imaging light paths do not overlap on the cornea and crystalline
lens surfaces. Pupil dilation is frequently required for wide field fundus
photography. Figure 5a2, b2 show trans-scleral illumination to achieve non-
mydriatic wide field fundus imaging. An optical fibre in contact with the sclera can
be used to deliver illumination light into the back of the eye10,11. the trans-scleral
illumination frees the entire pupil for imaging, and thus non-mydriatic wide field
fundus imaging is possible. However, the contact mode illumination is not
favourable, and thus it failed clinical deployment. We recently reported a prototype
of trans-palpebral illumination fundus camera, which delivered illumination light
through the eyelid, without direct contact to the sclera (Fig. 5a3, b3). However,
because of the absorption of the eyelid and sclera, the illumination light efficiency
was relatively low. To increase the illumination efficiency, we propose here to
develop contact-free trans-pars-planar illumination (Fig. 5a4, b4). As shown in Fig.
5a4, an arc-shaped illumination pattern, which matches the shape of the pars planar,
was projected to the pars planar without physical contact between the illuminator
and the sclera. After passing the pars planar, the illumination light was diffused and
illuminated the intraocular area homogenously (Fig. 5b4). Since corneal refection
was intrinsically eliminated and the entire pupil was used for imaging, wide field
fundus photography was made possible without pharmacologic pupil dilation (Fig.
5b4).

22. 16
Fig 4.2 Different image light path
4.2.3 Experimental Setup
Figure 6a shows the system diagram and Fig. 6b shows a photograph of the
lab prototype camera. A 565nm LED (Fig. 6c. M565L3, Thorlabs) was selected as
the light source for colour fundus imaging. Light from the LED was collected by a
lens and then passed through an arc-shaped aperture. A lens was used to image the
aperture onto the sclera to form an arc-shaped illumination pattern. the illumination
aperture was carefully designed to closely match the shape of the pars planar. the
end of the illuminating arm that was close to eye could be manually moved in a
horizontal direction by a translation stage to precisely deliver illumination light to
the pars planar. Light passing through the pars planar was diffused and illuminated
the intraocular area homogenously. A 22D ophthalmic lens (Volk Optical, Inc.) was
used to collect light coming out of the pupil. Tree of-the-shelf lenses (Thorlabs) were
placed after the ophthalmic lens to relay the fundus image onto the CMOS sensor of
a digital single-lens reflex camera (EOS Rebel T6i, Canon Inc.). An aperture was
placed at the pupil conjugate plane to restrict effective imaging pupil size to 2.5mm
for best imaging resolution24,25, as well as to reject scattering light from the sclera.
A lens was positioned behind the camera viewfinder and a cross that was illuminated
by an LED lamp was placed in front of the lens to serve as a fixation target, so that
the testing subjects could fix their eyes by looking into the camera through the lenses
and look at the cross. A single-shot fundus image could be easily acquired by
pressing the camera shutter button.
4.2.3.1 Fixation target: In conventional fundus cameras, a beam splitter can be used
to split light paths so that a fixation target could be implemented. However, a beam

23. 17
splitter wastes a fraction of the light from the retina. In our prototype system, there
was no beam splitter required due to the single reflex feature of the camera. As
shown in Fig. 6d, the reflex mirror reflected the light from the fixation target to the
eye so that the subject could see the Figure 5. Schematic illustration of different
illumination schemes for retinal imaging. (a1–a4) show illumination and imaging
light paths of trans-pupillary illumination (a1), trans-scleral illumination (a2), trans-
palpebral illumination (a3) and trans-pars-planar illumination (a4). (b1–b4) illustrate
available FOVs with trans-pupillary illumination (b1), trans-scleral illumination
(b2), trans-palpebral illumination (b3) and trans-pars-planar illumination (b4). Te T
shapes in b1 – b4 represent the pupil. When shutter was pressed, the reflex mirror
temporarily flipped up and light coming out from the eye reached the CMOS sensor
and formed a fundus image (Fig. 6d inset).
Fig 4.3 Mechanical setup
4.2.3.2 Human subject: This study was approved by the Institutional Review Board
of the University of Illinois at Chicago and was in compliance with the ethical
standards stated in the Declaration of Helsinki. Images shown were captured from
one healthy Asian female subject and one health Turkish male subject with informed
consent. No discomfort or vision impairment was reported by the subjects after
fundus photos being taken.
4.2.3.3 Light safety: Potential photochemical and thermal hazards of the retina were
carefully evaluated. There is no retina present at the pars planar area. However, it is
possible that the light passes through sclera and illuminates the retina. the thickness
of the sclera is ~0.5 mm26. the transmission of the sclera in visible wavelength is
10–30%27. To be conservative, 30% was used for calculation. According to the ISO

24. 18
15007-2:2007 standard, a maximum of 10 J/cm^2weighted irradiance is allowed on
the retina28 without photochemical hazard concern. The weighted irradiance was
calculated using the photochemical hazard weighting function provided in the
standard. For the proof-of-concept experiment, the weighted irradiance on the sclera
was calculated to be 0.5mW, the area of the arc-shaped light was 13 mm^2.For the
worst-case estimation, we assumed all illumination light directly expose to the
retinal area behind the illuminated sclera area. Therefore, the maximum allowed
exposure time is
T=
( / )
. ( )∗
(%)
( )
= 2.4s
If the illumination light accidently fell into the pupil, the illuminated area on retina
was estimated to be >9 mm^2.Thus the maximum allowed exposure time through
the pupil is >30minutes. For thermal hazard, the maximum weighted power intensity
allowed on the sclera without thermal hazard concern is 700mW/cm2 28. The
calculated weighted power intensity was 230mW/cm2, which was more than three
times lower than the maximum limit. Therefore, there was no thermal hazard
concern.
4.2.4 Result
Using all of-the-shelf components, we constructed the prototype camera for
proof-of-concept validation of trans-pars-planar illumination. Without the need of
pharmaType equation here.cologic pupil dilation, a 60° external-angle (90° eye-angle)
fundus view coverage was achieved in single-shot fundus images. All images were
red predominated because of the superior penetration capability of long (e.g., red)
wavelength light, compared to short (e.g., green and blue) wavelength light. For the
image, the average intensity of red channel was 4 and 16 times higher than that of
green and blue channels, respectively. In order to enhance the visualization of retinal
structures, red and green channels were digitally balanced. Given the absence of blue
light in the light source in the prototype instrument (see Methods: Experimental
setup), the blue channel was ignored to reconstruct the enhanced image. The macula
and optic disc were clearly observed, and individual blood
vessels were unambiguously identified. Moreover, nerve fibre bundles could also be
observed as stripped patterns coming from the optic disc.

25. 19
fig 4.4 output images
Fig 4.5 output images
Fig 4.6 schematic diagram

26. 20
4.3 DESIGN OF BELLOW ARRANGEMENT
Bellow is the mechanical setup that integrated with the android
mobile. This setup consists of optical lens system as well as illumination system.
This is manufactured by additive manufacturing technology. This is designed in
Solid works. This setup is designed based on android mobile dimensions. This
bellow setup also consists outer mobile holder case which is used to hold the mobile
phone.
4.4 IMAGING SENSOR
In our project the image sensor is android mobile phone camera.
Generally, there are two types of imagers i.e., CCD and CMOS. In android mobile
CMOS is used as light sensor. We also studied about the different image sensor
format available in CMOS.
Fig4.7 Block diagram of CMOS
After studying different image formats, we choose 1/2.55-inch image sensor i.e.,
12.2 megapixels is used for image capturing. We prefer Samsung mobiles with
1/2.55 inch to get good quality retinal images.

27. 21
Fig4.8 CMOS Table
Type 1/3.6" 1/3.2" 1/3" 1/2.7" ½.5" 1/2" 1/1.8" 1/1.7" 2/3" 1" 4/3"
Diagonal
(mm)
5.00 5.68 6.00 6.72 7.18 8.00 8.93 9.50 11.0 16.0 21.6
Width
(mm)
4.00 4.54 4.80 5.37 5.76 6.40 7.18 7.60 8.80 12.8 17.3
Height
(mm)
3.00 3.42 3.60 4.04 4.29 4.80 5.32 5.70 6.60 9.6 13.0
Area
(mm2
)
12.0 15.5 17.3 21.7 24.7 30.7 38.2 43.3 58.1 123 225
Fig4.9 Image sensor format

28. 22
CHAPTER 5
SOFTWARE DESCRIPTION
5.1 ZEMAX OPTICSTUDIO
Optic Studio is an optical design program that is used to design and
analyse imaging systems such as camera lenses, as well as illumination systems. It
works by ray tracing modelling the propagation of rays through an optical system. It
can model the effect of optical elements such as simple lenses, aspheric
lenses, gradient-index lenses, mirrors, and diffractive optical elements, and can
produce standard analysis diagrams such as spot diagrams and ray-fan plots. Optic
Studio can also model the effect of optical coatings on the surfaces of components.
It includes a library of stock commercial lenses. Optic Studio can perform standard
sequential ray tracing through optical elements, non-sequential ray tracing for
analysis of stray light, and physical optics beam propagation. It also has tolerancing
capability, to allow analysis of the effect of manufacturing defects and assembly
errors.
The physical optics propagation feature can be used for problems
where diffraction is important, including the propagation of laser beams and the
coupling of light into single-mode optical fibres. Optic Studio’s optimization tools
can be used to improve an initial lens design by automatically adjusting parameters
to maximize performance and reduce aberrations.

29. 23
5.1.1 Zemax optic studio Programming Windows
Fig 5.1 Editor Window
5.1.2 Lens data editor
Fig 5.2 Lens data

30. 24
5.2 IMAGE SIMULATION OUTPUT
In optic studio we can simulate image of the object using different
bitmap images which are inbuilt in the software libraries. We can also import
images in the library. We can obtain spot diagram of output image. There are
different types of image simulation like geometric image analysis, partially
coherent image analysis, extended diffraction analysis is available in this
software.
Fig 5.3 Output Image analysis

31. 25
Fig 5.4 Spot Diagram
Fig 5.5 3D lens setup

32. 26
CHAPTER 6
PROJECT DESCRIPTION
6.1 WORKING PRINCIPLE
Fundus photography is used to take both retinal and cornea images of
human eye. This works on the principle of indirect ophthalmoscopy. The optical
lens system in the device captures retinal and cornea images with the help of
illumination system. The illumination system in the device is obtained by palpebral
illumination technology in which the warm white light is placed at 90 degree to the
sclera i.e. above the eye. This provide lighting inside the eye which helps for image
capturing. The captured image by lens system is detected by imaging sensor i.e.
through android mobile phone camera and displayed on phone screen. The image
undergoes image processing and stores the output data in the android application.
The different views of the patient eye are used for eye diagnosis and determine the
eye diseases like cataract eye, diabetics retinopathy etc. Thus the smart phone
based fundus camera is used for mass eye screening. This device works in non-
mydriatic condition in which there is not necessity of eye pupil dilation to increase
the pupil diameter.
6.1.1 Image Capturing
It is the first part of the system in which the retinal and cornea images are
captured using optical lens system. As it is non mydriatic device there is no need any
dilation before capturing image. To capture the image we should place the fundus
device before the patient eye at a distance of 35mm. We should also take care about
alignment of mobile camera. The centre of the primary camera should be in co-axial
with the centre of pupil. We should also place illumination setup on the eye lid
carefully. The intensity of the warm light is electronically controlled.
When the illumination system is on, the image sensor captured the retinal and
cornea images using optical lens system. The images are captured in different angles.
6.1.2 Image Processing Phase
The output image of the image sensor has some geometric aberrations. To
eliminate the aberrations and obtained sharp focused image, the output image

33. 27
undergoes image processing in the application developed by software team. After
pre-processing and enhancing images is stored as a zip file in android application.
6.1.3 Diagnosis
We can transmit patient data through cloud or providing a login id and
password to the doctor. Doctor can login into the android application through login
details and go through the patient data. In the application itself the right and left eye
images are classified. Doctor can go through the images and diagnose the condition
of the patient eye and give treatment if there is any problem in the eye.
6.2 PROJECT DESIGN
Fig 6.1 Block diagram
6.3 PROCEDURE
STEP 1: switch on the illumination system and align the fundus camera before eye
STEP 2: Capture the retinal image of the eye using mobile camera
STEP 3: The captured image is pre-processed and enhanced
STEP 4: The processed image is stored in the android application
STEP 5: Output images are classified in the software
STEP 6: The patient data is sent to Ophthalmologist.
illumination
setup
patient eye
output lens
system
image sensor

34. 28
STEP 7: The doctor examines the retinal and cornea images and diagnose the
patient.
6.4 CALCULATION AND MEASUREMENT
Here we consider the dimensions of emmetropic eye.
● Effective focal length of eye model is 17mm
● Effective power of eye is 60 dioptres the first set of lenses in eyes
● Total object height = 24mm
● Observable height = 18mm
● Magnification = 3.13
● Keeping a constrain of 45 degrees field of view we can calculate the working
distance, using the following formula
𝐴𝑛𝑔𝑢𝑙𝑎𝑟 𝐹𝑖𝑒𝑙𝑑 𝑜𝑓 𝑉𝑖𝑒𝑤(𝐴𝐹𝑂𝑉) =
𝑐𝑜𝑛𝑑𝑒𝑛𝑑𝑖𝑛𝑔 𝑙𝑒𝑛𝑠 𝑑𝑖𝑎𝑚𝑒𝑡𝑒𝑟
𝑝𝑜𝑤𝑒𝑟 𝑜𝑓 𝑙𝑒𝑛𝑠 × 𝑤𝑜𝑟𝑘𝑖𝑛𝑔 𝑑𝑖𝑠𝑡𝑎𝑛𝑐𝑒(𝑊)
Note 1: We cannot directly use the formula because the light passes
through different fluids having dissimilar refractive index
Note 2: The working distance is calculated from cornea and if we make
the projection of pupil on the surface of cornea it makes a circle with diameter
ranging from 2.1mm to 6mm.light will be traveling through this small pupil
orifice. the retina is placed at a distance of about 24mm from the inner surface
of cornea. The light converges at a rate of 60D and this occurs for a distance
near to 13mm and after passing this zone it enters the eye cavity having
refractive index of 1.34 slightly greater than air. all these factors induce a
convex behaviour causing the ray to converge.
Working distance (w) = 47mm from cornea

35. 29
6.5 OUTPUT IMAGES
Fig 6.2 Retinal images in different views
Fig 6.3 Eye Dimensions

36. 30
Fig 6.4 comparison Chart
6.6 EXPERIMENTAL SETUP
Fig 6.6 Experimental setup

37. 31
CHAPTER 7
CONCLUSION
The smart Phone based fundus camera is designed and manufactured for
capturing retinal and cornea images successfully. This system is used for mass
screening in rural area. This is an active device and also a cost efficient which is
easily affordable. This device is work as non- mydriatic condition which reduces the
irritation of patient. This can be easily handled by anyone so screening can done by
lab technician and need not require any ophthalmologist. This has more advance
features when compared to other devices available in the market. This can further
modified in lens design to capture more focused images. We can also develop a
software program that identifies the retinal diseases by using machine learning
technology in future.

38. 32
REFERENCES
[1]. Mann WA. History of photography of the eye. Sur Ophthalmol. 1970;15:179–
189.
[2]. Van Cader TC. History of ophthalmic photography. J Ophthalmic Photogr.
1978;17–19.
[3]. Bennett TJ, Barry CJ. Ophthalmic imaging today: an ophthalmic
photographer’s viewpoint—a review. Clin Experiment Ophthalmol. 2009;37:2–13.
[4]. Gutner R, Miller D. Inside the fundus camera. Ann Ophthalmol. 1983;15:13–
16.
[5]. DeHoog E, Schwiegerling J. Fundus camera systems: a comparative analysis.
Appl Opt. 2009;48:221–228.
[6]. Gliss C, Parel JM, Flynn JT, Pratisto H, Niederer P. Toward a miniaturized
fundus camera. J Biomed Opt. 2004;9:126–131.
[7]. Yogesan K, Constable IJ, Barry CJ, Eikelboom RH, McAllister IL, Tay-
Kearney ML. Telemedicine screening of diabetic retinopathy using a hand-held
fundus camera. Telemed J. 2000;6:219– 223. 7606 Tran et al. IOVS, November
2012, Vol. 53, No. 12 D 09/09/2018
[8]. Cheung N, Mitchell P, Wong TY. Diabetic retinopathy. Lancet.
2010;376:124–136.
[9]. Mohamed Q, Gillies MC, Wong TY. Management of diabetic retinopathy: a
systematic review. JAMA. 2007;298:902–916.
[10]. Saligan LN. Preventing diabetic retinopathy in primary care. Nurse Pract.
2008;33:46–47.