Ihtatho, Dani (2008) Objective Assessment of Area and Erythema of Psoriasis Lesion Using Digital Imaging and Colourimetry. Masters thesis, Universiti Teknologi Petronas.
2008 Master - Objective Assessment of Area and Erythema of Psoriasis Lesion Using Digital Imaging.pdf
Download (8MB)
Abstract
Psoriasis is a non-contagious skin disease which typically consists of red plaques covered
by silvery-white scales. It affects about 3% of world population. During treatment,
dermatologists monitor the extent of psoriasis continuously to ascertain treatment
efficacy. Psoriasis Area and Severity Index (PAS!) is the current gold standard method
used to assess the extent of psoriasis. In PAS!, there are four parameters to be scored i.e.,
the surface area affected, erythema (redness), thickness and scaliness of the plaques.
Determining PAS! score is a tedious task and thus it is not used in daily clinical practice.
In addition, the PAS! parameters are visually determined and may result in intra-observer
and inter-observer variations, even by experienced dermatologists. Objective methods in
assessing area and erythema of psoriasis lesion have been developed in this thesis.
Psoriasis lesion can be recognized by its colour dissimilarity with normal skin. Colour
dissimilarity is represented by colour difference in CIELAB colour space, a widely used
colour space to measure colour dissimilarity. Each pixel in CIELAB colour space can be
represented by its lightness (L'), hue (hob), and chroma (Cab). Colour difference between
psoriasis lesion and normal skin is analyzed in hue-chroma plane of CIELAB colour
space. Centroids of normal skin and lesion in hue-chroma space are obtained from
selected samples. Euclidean distances between all pixels with these two centroids are
then calculated. Each pixel is assigned to the class of the nearest centroid. The erythema
of psoriasis lesion is affected by degree of severity and skin pigmentation. In order to
assess the erythema objectively, patients are grouped according to their skin pigmentation
level. The L* value of normal skin which represents skin pigmentation level is utilized to
group the patient into the three skin types namely fair, brown and dark skin types. Light
difference (t.L*), hue difference (t.hab), and chroma difference (t.C'ab) of CIELAB
colour space between reference lesions and the surrounding normal skin are analyzed. It
is found that the erythema score of a lesion can be determined by their hue difference
(t.hab) value within a particular skin type group. Out of 30 body regions, the proposed
method is able to give the same PAS! area score as reference for 28 body regions. The
proposed method is able to determine PAS! erythema score of 82 lesions obtained from
22 patients objectively without being influenced by other characteristic of the lesion such
as area, pattern, and boundary.
Item Type: | Thesis (Masters) |
---|---|
Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering |
Departments / MOR / COE: | Engineering > Electrical and Electronic |
Depositing User: | Users 2053 not found. |
Date Deposited: | 09 Oct 2013 11:07 |
Last Modified: | 15 May 2023 03:56 |
URI: | http://utpedia.utp.edu.my/id/eprint/8543 |