Use this URL to cite or link to this record in EThOS:
Title: Simulating the colour of port wine stain skin
Author: Lister, Thomas
ISNI:       0000 0004 2734 4858
Awarding Body: University of Southampton
Current Institution: University of Southampton
Date of Award: 2013
Availability of Full Text:
Access from EThOS:
Access from Institution:
Currently, laser treatments for Port Wine Stain (PWS) lesions are considered the choice therapy, but response is poor or treatments are ineffective for around half of patients. It is proposed in this thesis that improvements to the effectiveness of laser treatment can be achieved through the acquisition of estimated PWS vessel number density, depths and diameters for each individual lesion. Information regarding PWS vessel architecture is found to be contained within the colour of the lesion. Presented in this thesis is a method of extracting this information through colour measurements and the inverse application of a skin model. Colour measurements are performed on 14 participants using a Konica-Minolta CM2600d spectrophotometer employing a xenon flashlamp illumination source and an integrating sphere. Light transport is simulated through an 8 layer mathematical skin model inclusive of horizontal, pseudo-cylindrical PWS blood vessels using a new Monte Carlo programme. Within the programme, model parameters were adjusted in an iterative process and skin colour was reproduced with a mean discrepancy of 1.9% reflection for clinically normal skin (24 datasets) and 2.4% for PWS skin (25 datasets). The programme estimated anatomical properties of the measured regions of skin, yielding epidermal melanin volume fractions from 0.4% to 3.3% and mean melanosome diameters from 41 nm to 384 nm across the participant group. The response to laser treatment was assessed for 10 participants through colour measurements taken immediately before and at least 6 weeks after treatment and through expert analysis of photographs for 9 participants taken at these times. Treatment response was not found to correlate directly with the pre-treatment melanin parameters estimated by the programme. Mean depths, diameters and number densities of PWS vessels were also estimated by the programme before and after treatment. These parameters were compared to data obtained from Optical Coherence Tomography (OCT) images for 5 participants. Number densities and diameters predicted by the simulation varied by no more than 10% from the values determined by OCT for 4 and 5 out of 7 regions respectively. Mean depths predicted by the simulation did not correspond with those determined by OCT however. This may be a result of the limited contribution of deeper vessels to the colour of PWS skin. Predicted PWS parameters were compared to treatment response assessed by colour measurement for 10 participants and by photographic analysis for 9 of these. Predicted vessel number densities were not found to correspond with treatment response. Vessel diameters predicted by the simulation correlated with treatment response when compared with the pulse lengths selected for treatment. Optical coefficients derived from the skin model were used to estimate appropriate laser treatment radiant exposures at the predicted mean vessel depths and these radiant exposures corresponded strongly with the treatment response. Suggestions for improvements in the predictions of melanosome diameters through changes in the adjacent skin minimisation procedure within the programme are discussed. The apparent underestimation of PWS blood vessel number densities and mean depths (compared to biopsy studies) may be a result of the reduced influence of deeper PWS vessels upon skin colour. Further investigation, including modifications to the PWS vessel minimisation procedure within the programme, would be necessary to determine whether improvements in these predictions may be achievable. The results of the study show that the new Monte Carlo programme is capable of extracting, from measurements of skin colour, realistic estimates of PWS skin characteristics which can be used to predict treatment response and therefore inform treatment parameters on an individual PWS.
Supervisor: Chappell, Paul Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: QA75 Electronic computers. Computer science