Mobile Application for Ulcer Detection
be recorded using cameras that can capture
mid wave or long wave infrared light. The
latter is cheaper and can cover wider ranges
of temperature. The mid–wave infrared
thermal cameras are more accurate and
therefore better used for medical purposes [15] .
However, for home use, the long wave infrared
thermal camera can be considered perfect.
An example of the long wave infrared thermal
camera is the FLIR ONE [16] . Studies have been
conducted on preventive techniques for DFU
but without using an android compatible FLIR
ONE device. Those studies comprise of long,
and non-user friendly methods of detection.
Color images were used to determine the
presence of ulcers and its healing process,
but the ulcers would already have to be
at their advanced stages [17] . Temperature,
humidity, and pressure were measured, by
another study, across the soles of the foot by
carefully placing the sensors using a smart
shoe to determine foot inflammation, which
is an indicator for DFU [18] . The sensors send
data to the mobile application via Bluetooth
and a server helps to process the data. A
study proposed the combination of digital
photography with infrared thermography
to acquire the color image and infrared
thermometer [7] . FLIR SC305 handheld device
was used in some studies to study the early
detection of DFU; comprising of a huge
apparatus that is not portable, and the images
had to be exported to a computer for analysis,
run on a server based system from which the
results are obtained – making it a long and
tedious process [3, 19] . Another study discussed
various techniques of DFU prevention
using infrared thermal camera [20] . Robust
acquisition protocol for early preventive
measures for DFU detection has also been
proposed as a viable solution [13] . Studies verify
that a temperature difference of 2.2 degree
Celsius or 4 degree Fahrenheit is a clear
indication for the presence of DFU in diabetic
patients [3, 21, 22] . Another feasibility study
conducted by Fraiwan et al. [26] for a Matlab
mobile detection system utilizes a Matlab
based approach to the early detection of DFU
problem. The difference between the current
study and this study is the methodology and
the outcome of both studies. The current study
aimed at implementing a standalone mobile
application for the detection of diabetic foot
ulcers under simulated patient conditions. It
was completely implemented using Android
Studio, open CV image processing Library,
and FLIR ONE SDK. The application was a
standalone system.
Today, technologies exist that can help
determine the presence of ulcers in a diabetic
foot. Most of these technologies, however, are
large, non-portable, expensive, sometimes
invasive, and require an expert physician. This
work aims at constructing a standalone Java
based mobile thermal imaging system that
can be used by diabetic patients to identify
possible ulcerous regions in their feet or any
other location.
"Most of these technologies, however, are
large, non-portable, expensive, sometimes
invasive, and require an expert physician."
Current Pedorthics | January/February 2020
19