Quantification of prognostic parameters for assessment of diabetes-related foot ulcers and venous leg ulcers using image processing techniques

Rani, P 2019, Quantification of prognostic parameters for assessment of diabetes-related foot ulcers and venous leg ulcers using image processing techniques, Doctor of Philosophy (PhD), Engineering, RMIT University.


Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Rani.pdf Thesis application/pdf 2.62MB
Title Quantification of prognostic parameters for assessment of diabetes-related foot ulcers and venous leg ulcers using image processing techniques
Author(s) Rani, P
Year 2019
Abstract Diabetes-related Foot Ulcers (DFUs) and Venous Leg Ulcers (VLUs) are two important types of chronic wounds which do not heal in an orderly fashion and being a major cause of morbidity in extreme cases. Assessment of these ulcers is a growing concern among health care professionals across the globe. Currently, healing of these ulcers is assessed by monitoring the changes in their area over four consecutive weeks. The suggested clinical monitoring parameter is that the ulcers which show more than fifty percent reduction in the area by week four (after the ulcers are reported in the clinics) are predicted to heal within twelve weeks of time. However, this is a subjective measurement, performed manually using a ruler; based on the assumption that the ulcer is purely rectangular. Moreover, the above-mentioned monitoring method fails to work in most of the cases, as healing of these ulcers is a complicated, multifactorial process which cannot be assessed only by a single parameter (i.e., area).

This research work has proposed new objective parameters for assessment and prediction of these ulcers by studying the shape of the ulcers, temperature distribution of the ulcer and on the ulcerated foot and area measurement of the ulcers using different techniques. This work has also examined the association of the proposed parameters with patient's clinical information, etiological factors and the healing status of the ulcers. Literature has suggested that there is a change in the irregularity of the ulcers as they heal, thus, playing the role in the healing of the ulcers. Based on this fact, this work hypothesized that the edges of the ulcers can be assessed by quantifying the change in irregularity in them. The widely used technique of measurement of irregularity is Fractal Dimension (FD). However, FD has the limitation of inherent limited resolution of the digitized images, which renders these images as non-fractal, as the self-similarity properties of the images are lost. Thus, this work proposed a new index measure and developed an algorithm for measurement of irregularity and tested it on synthetic images initially. The new index measure called curve irregularity index (Ic) measures the change in the irregularity of the segments of the contours with change in window sizes and does not assume the objects to have self-similar properties. The Ic was then measured and validated on the contours of DFUs and VLUs and the association of irregularity of the contours with etiological factors and the healed status of the ulcers, respectively, was reported. This work has employed the normal DSLR camera and digital planimetry technique to capture the RGB images and to obtain the tracings of the ulcers respectively. This work has shown the significance of contour irregularity of ulcers with the clinical conditions of patients and differentiated between the healed and not-healed ulcers.

This research has also employed infrared thermal imaging technique to obtain the temperature distribution of the ulcers. Literature has reported the association of temperature with the risk of ulceration in neuropathic and ischemic conditions of the feet. However, very few works have been done on the study of temperature of the existing ulcers. Hence, this work tested and obtained the association of the mean temperature of the DFUs with the clinical conditions of the participants. This work also hypothesized that the area obtained based on thermal distribution can differentiate between the healed and not-healed ulcers in VLUs. Hence, segmentation of the ulcer region from the thermal images was performed based on an active contour model, previously developed for segmentation of contours in images where edges are not defined by gradient. The obtained results showed that the area thus obtained from the ulcer regions of all five weeks showed association with the healed status of the VLUs and can differentiate between the healed and not-healed ulcers. This work can also be used to predict the healing trajectory of the ulcers. Thus, the overall research work would find application in the clinical set ups to aid in the assessment and prediction of the healing status of DFUs and VLUs and would lead to provide better health-related quality of lives to the patients.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Engineering
Subjects Biomedical Instrumentation
Keyword(s) diabetic foot ulcers
venous leg ulcers
thermal imaging
digital planimetry
image processing
Versions
Version Filter Type
Access Statistics: 122 Abstract Views, 72 File Downloads  -  Detailed Statistics
Created: Mon, 11 Nov 2019, 12:18:19 EST by Keely Chapman
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us