1.

Objectives:

1- Primary Objectives:
 To develop an AI-based handheld fundus camera that achieves diagnostic
accuracy greater than 90% in detecting common retinal diseases (e.g., diabetic
retinopathy, glaucoma), comparable to standard clinical fundus cameras.

 To Reduce the cost of fundus imaging devices by at least 60%

Compared to conventional market devices priced between USD 10,000–50,000.

 Increase access to retinal screening services by at least 40%

In underserved and remote areas and primary healthcare centers.

 Achieve an operating profit margin of at least 50–60%

To ensure financial sustainability, scalability, and commercial viability of the
product.

 Reduce examination and diagnosis time by 50%

Through real-time image analysis and automated AI-based reporting.

2- Secondary Objectives:
 To design a portable, lightweight, and ergonomically optimized handheld fundus
camera suitable for use in low-resource environments.

 To enable AI-driven real-time image analysis and automated diagnostic reporting at

the point of care.

 To ensure high image quality and consistency comparable to standard tabletop

fundus cameras.

 To reduce examination time per patient by minimizing manual interpretation and

referral delays.

 To lower the cost of retinal screening through affordable hardware design and
automated diagnostics.

 To support tele-ophthalmology and remote consultation via secure image storage

and transmission.
Mission and Values
Our Mission
Our mission is to prevent avoidable blindness by improving access to reliable and
affordable eye screening services through locally developed technology that supports the
healthcare system and community needs.

Our Vision
We aim to become a leading local provider of handheld fundus cameras and intelligent
screening solutions for blinding eye diseases, with a strong focus on serving clinics, primary
healthcare centers, and underserved regions.

SWOT Analysis for AI FundusCam:

Strengths

 Portable and lightweight design enables use in remote and primary healthcare settings
 AI-based image analysis allows rapid and automated detection of retinal diseases
 Lower cost compared to traditional tabletop fundus cameras
 Reduces dependency on specialist ophthalmologists for initial screening
 Supports early detection of common retinal diseases (e.g., diabetic retinopathy,
glaucoma)
 User-friendly interface suitable for non-specialist healthcare workers

Weaknesses

 Image quality may be lower than high-end stationary fundus cameras

 AI accuracy depends heavily on the size and diversity of training datasets
 Requires regular software updates and model validation
 Limited battery life for prolonged field use
 Initial regulatory approval and clinical validation can be time-consuming
 Dependence on stable lighting and proper user handling for optimal results
Opportunities

 Growing global demand for affordable eye screening solutions

 Increasing prevalence of diabetes and age-related eye diseases
 Strong potential for deployment in underserved and rural regions
 Integration with telemedicine and electronic health record (EHR) systems

Threats

 Competition from established medical imaging device manufacturers

 Regulatory changes affecting AI-based medical devices
 Data privacy and cybersecurity concerns related to patient information
 Risk of clinician resistance to AI-assisted diagnosis
 Rapid technological advancements that may shorten product lifecycle
 Potential liability issues related to AI misdiagnosis

Key To succuss:

Clear Medical Problem Identification

The device addresses a real healthcare need, such as:
Limited access to fundus imaging devices in remote areas.
High cost of traditional fundus cameras.
The need for early detection of retinal diseases (e.g., diabetic retinopathy, glaucoma).

Simple and User-Friendly Design

Lightweight and portable design.
Easy to use by doctors or healthcare workers without complex training.
Suitable for small clinics and rural healthcare centers.

High Image Quality and Accuracy

Clear and clinically usable retinal images.
Reduced noise and improved illumination.
Auto-focus and image enhancement features.
 Cost-Effectiveness Compared to Competitors
Significantly lower cost than traditional devices.
Reduced maintenance and replacement costs.
Affordable for low-resource healthcare settings.

Smart Connectivity and Data Management

Secure storage of patient records.
Ability to share images remotely for telemedicine purposes.

3.0:
Product Description

Description Item

AI FundusCam Product Name

Image

- Portable and lightweight design

- High-resolution retinal imaging
- Auto-focus and illumination control Specifications
- Battery operated (long-lasting)
- Stores patient data securely

- Early detection of eye diseases

- Easy to use in clinics and remote areas
- Affordable alternative to traditional devices Customer Benefits
- Quick image capture and sharing for telemedicine
- Reduces patient waiting time and improves diagnosis accuracy
 Opportunity for
Gaps (What is Weaknesses Strengths (What
Our Handheld AI Company
missing) (Limitations) they do well)
Fundus Camera
• Provide a low-cost,
• No low-cost • Very strong
handheld device • • Very expensive
handheld solution • brand in eye care •
Use AI for fast devices • Large,
Limited focus on Excellent image
screening, not full non-portable
rural or primary quality • Widely
diagnosis • Enable systems • Requires Topcon
care screening • AI used in hospitals
use by nurses and trained specialists •
not optimized for and eye clinics •
technicians • Reach Not designed for
simple, fast Advanced clinical
rural and remote field use
screening systems
areas
• Build an offline AI
• No strong AI- • Less focused on
system • Focus on • High-quality
driven screening eye-care
retinal disease imaging sensors •
device • AI depends specialization •
screening • Serve Reliable and easy
on internet/cloud • Limited built-in AI Canon
clinics without to use • Strong
Not suitable for diagnosis • Mostly
internet • Target global production
low-connectivity clinic-based
primary healthcare and support
areas equipment
centers
• No true handheld • Good clinical
• Introduce AI-first • Limited AI
AI device • No imaging quality •
handheld camera • features • Slow
automated More affordable
Automatic disease innovation in
screening workflow than Topcon • Nidek
detection and reports portable devices •
• Weak presence in Comfortable and
• Support mass Focused mainly on
community simple for
screening programs clinics
screening technicians

Opportunties:

 AI provides first (initial) screening only

 Final diagnosis by ophthalmologist

 Handheld and portable device

 Low cost compared to existing systems

 Usable by non-specialists

 Instant screening results

 Works without internet

 Designed for primary care and rural areas

 Reduces burden on eye specialists

 Enables early detection of retinal diseases

 Sourcing :

Benefits to the Why We Chose Supplied

Who They Are Supplier
Project Them Components
• Short lead time • • Easy local access • Power Local Al-Haddad
 Benefits to the Why We Chose Supplied
Who They Are Supplier
Project Them Components
Lower procurement
Components, electronics
cost • Quick
Fast availability • voltage supplier
replacement and
Competitive local regulators, specializing in Electronics
support • Reliable
pricing cables, power
supply for power
connectors components
systems
• Trusted supplier of
• High optical clarity • Specialized
medical-grade
Reliable and supplier of Ibn Al-
optical lenses • 20D ophthalmic
consistent lens quality ophthalmic Haytham
Suitable optical lens
• Supports accurate optics and Company
quality for fundus
retinal image capture lenses
imaging
• Accelerates
prototyping and • Provides
testing • Reduces components not Raspberry Pi,
Electronics and Al-Qariya
international shipping available locally • touch screens,
scientific (Scientific
delays • Ensures Reliable regional embedded
equipment Village Store,
availability of supplier • Faster computing
supplier Saudi Arabia)
specialized delivery than modules
development overseas sourcing
components
• Immediate
• Quick access to Local
availability of imaging
compatible camera technology
hardware • Local Hikvision Local Camera
hardware • Supports store selling
warranty and support • camera modules Store
prototype camera
Easy testing and
development modules
integration

Hardware Technology –
 Technology /
Benefit Reason for Use
Component

Provides clear images for early disease To capture high-resolution

Camera Module
diagnosis retinal images

Manages data, camera operation, and image

Processing and control unit Raspberry Pi Board
processing efficiently

Proper magnification, wide field of view, ease To obtain detailed and clear 20D Retinal Imaging
of use for both patient and clinician retinal images Lens

Enhances image clarity and diagnostic Provides proper retina LED / Xenon-Halogen
accuracy with accurate color rendering illumination during imaging Light Source

Patient registration, image display, ease of User interface for device 3.5 Inch LCD
use interaction Touchscreen

Maintains stable Raspberry Pi

Ensures safe and reliable device operation Cooling Fan
temperature

Supplies power to all Power Supply /

Continuous and stable operation
components Charger

High-quality display, responsive interaction, Connects Raspberry Pi to the

HDMI Cable
clear image output display

Easy transport and usage in clinics or remote Provides a portable, protected Enclosure / Portable
health centers device Case

Software Technology –

Benefit Reason for Use Technology

Python
Simple, powerful, supported by extensive Core programming language
Programming
libraries, suitable for image processing and AI for system development
Language

Image resizing, brightness/contrast

Image processing and OpenCV / Pillow
adjustment, format conversion, enhancing
preparation for AI (PIL)
image quality

Creating charts and graphs for performance

Data visualization and analysis Matplotlib
analysis

Retinal image analysis for early disease

Building and training AI models TensorFlow / Keras
detection using deep learning

Generating classification reports, confusion Model evaluation Scikit-learn

matrices, ROC-AUC, measuring model
 Benefit Reason for Use Technology

performance

Automated loading and organization of images File and dataset path

OS Library (Python)
efficiently management

User-friendly interface for patient registration, Creating graphical user

Tkinter (GUI Library)
image display, and examination interface

Structured, secure, and retrievable data Storing patient data and

MySQL Database
management images

Pre-trained deep learning

High accuracy, reduces training cost, fast and VGG16 Pre-trained
model for retinal disease
reliable for early detection CNN Model
detection

Future Work

1. Device Improvement: Non-Mydriatic Functionality

o Transform the current mydriatic fundus camera into a non-mydriatic device,
allowing retinal imaging without pupil dilation.
o Enhance patient comfort and reduce preparation time during screenings.
2. AI Model Enhancement
o Develop an AI model capable of learning from experience to improve
accuracy over time.
o Expand training datasets to cover diverse populations and rare retinal
conditions.
o Incorporate adaptive learning for continuous performance improvement.
3. Hardware Upgrades
o Integrate higher-resolution cameras for sharper retinal images and better
diagnostic capability.
o Improve battery efficiency to support longer screening sessions.
4. Software and User Interface
o Implement real-time image quality checks and guidance for operators.
o Maintain user-friendly interface suitable for non-specialist users.
o Enable multi-language support to expand accessibility.
5. Connectivity and Telemedicine
o Provide options for remote specialist review via secure cloud or local servers.
o Support automated report generation for easy integration with medical records.
6. Clinical Validation and Regulatory Compliance
o Conduct clinical trials to validate non-mydriatic AI-assisted screening
accuracy.
o Pursue necessary approvals (e.g., CE, FDA) for medical deployment.