Transcript PathoCam: Point-of-Care Medicine with an Android Camera Phone

PathoCam: Point-of-Care Medicine with an
Android Camera Phone
WAN QI CHOO and JUSTIN SOHN
School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN 47907
Advisors: JAN P. ALLEBACH, EDWARD J. DELP, RON REIFENBERGER
A VIP Project – Spring Semester 2013
A GLOBAL NEED – PERSONALIZED
MEDICAL CARE
CAMERA PHONES - A WAY FORWARD?
The impressive penetration of mobile cellular phones worldwide
is driving rapid advances in telemedicine.
Estimated 4.6 billion cell phones have been sold worldwide.
• Medical diagnostics comprise <5% of total
hospital costs, yet they influence 60-70% of
healthcare decisions.
• In developing nations, expenditure on medical diagnostics is minimal. There is little interest in financing
diagnostic development due to perceived lack of return on investment.
• Highly sensitive tests are expensive and therefore inaccessible at remote point-of-care locations.
QUESTION: Could camera phones be used to meet the
increasing demand for inexpensive point-of-care medical tests
that are easily administered by an untrained patient?
GOAL: Develop a point-of-care medical test that relies on a
dedicated ‘cell phone app’.
IMPACT: Health care world-wide could be revolutionized.
• An innovative, disruptive, and inexpensive paradigm shift in diagnostic technology is required.
PROJECT BACKGROUND – PATHOGEN IDENTIFICATION BY PATTERN RECOGNITION
At Purdue an interdisciplinary team (Phil Low and Alex Wei, Chemistry; Yeong Kim and Ron Reifenberger, Physics)
has developed a novel approach to point-of-care bacterial identification. Chips patterned with immutable ligands
capture targeted bacteria into pre-determined patterns that can be identified remotely using pattern recognition
algorithms. The technique is fault-tolerant, robust and simple, requiring only the analysis of camera phone images of
the chip after exposure. The five-step approach is summarized below:
PATHOCAM
'APP '
1. Distribute patterned chip
5. Central location
for pattern
analysis. Results
returned directly to
point of origin.
3. Captured
bacteria forms
pattern
Application
Development
Characteristics:
•Easy to program
•Widely Available
•Useful Features
•Future Updates
http://www.statisticbrain.com/android-phone-statistics/
Application Features:
4. Cell Phone Imaging
2. Expose to test fluid
(15 mins.)
Selecting a Cellular Phone
•Easy to use
•Ability to capture an image
•Ability to send image to remote location
•Receive results
SAMSUNG Android
Selected
Current Development
http://www.moneyweb.co.za/moneyweb-technologynews/most-popular-cellphone-brands-in-the-world
Looking To The Future
Two-Way Communication
The next goal is to create a server that
will automatically identify the illness
and report it to the user.
Environment:
•Android development tool:
Eclipse
.
GPS Coordination
•Programming language:
JAVA
•Android compatibility:
2.2 – 4.2
Most Popular Mobile Phone
Brand by Gartner Inc.
To map the disease so that the user
can see the distribution of the infection
Home Screen
Capture Chip
Image
Confirm
Image
Input User
Email
Address
Receive
Feedback
Image Saved
in PathoCam
Folder
Use Unique
Filename:
IMEI_DateTime
.jpg
Special thanks to Shao-Fu Xue and Albert
Parra-Pozo for all their help