Muhammad Imran Iqbal; Abiodun Musa Aibinu; Nagendra Srinivas Gubbal; Asim Khan , pp. 62. TEK/avd. för signalbehandling, 2006.
This thesis applies the process and knowledge of digital signal processing and image
processing to diagnose diabetic retinopathy from images of retina.
The Pre-Processing stage equalizes the uneven illumination associated with fundus
images and also removes noise present in the image. Segmentation stage clusters
the image into two distinct classes while the Disease Classifier stage was used to
distinguish between candidate lesions and other information. Method of diagnosis of
red spots, bleeding and detection of vein-artery crossover points were also
developed in this work using the colour information, shape, size, object length to
breadth ration as contained in the digital fundus image in the detection of this
In addition to diagnosis of Diabetic Retinopathy (DR), two graphical user interfaces
(GUI’s) were also developed during this work, this first is for collection of lesion data
information and was used by the ophthalmologist in marking images for database
while the second GUI is for automatic diagnosing and displaying the diagnosis result
in a more friendly user interface and is as shown in chapter three of this report.
The algorithm was tested with a separate set of 25 fundus images. From this, the
Receiver Operating Characteristics (ROC) was determined for red spot disease and
bleeding, while cross over points were only detected leaving further classification as
part of future work needed to complete this global project. Sensitivity (classify
abnormal fundus images as abnormal) and specificity (classify normal fundus image
as normal) was calculated for the algorithm is given as 98% and 61%.