SUBSURFACE IMAGING USING CROSS POLARIZED LIGHT FOR ENHANCED CLINICAL CHARACTERISTICS VISUALIZATIONS AND QUANTIFICATION OF ACNE SKIN ON THE FACE IMAGE
Acne, or acne vulgaris, is one of the most common disease that we encounter. Almost 85% of adolencents have experienced this disease, especially on the face. In dermatology, medical imaging for observation and clinical evaluation that emphasize objective assessment is needed because usually the asse...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Subjects: | |
| Online Access: | https://digilib.itb.ac.id/gdl/view/75410 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Acne, or acne vulgaris, is one of the most common disease that we encounter. Almost 85% of adolencents have experienced this disease, especially on the face. In dermatology, medical imaging for observation and clinical evaluation that emphasize objective assessment is needed because usually the assessment depends solely on subjective opinion of the doctor. For the case of acne, the characteristics images that are our primary concern clinically are erythema and inflammatory lesion. A usual medical imaging arrangemets are unable to capture characteristics of an acne. A configuration of imaging device using cross polarized light emphasizes imaging of an subsurface layer of the skin so that it could capture clinical characteristics of an acne skin. This device are able to reject the glare or light that reflected from skin surface and accepts light that diffusely reflected from subsurface tissue and depolarized by dermal collagen tissue birefringent.
In this final project research, cross polarized subsurface imaging configuration will be constructed for capturing acne skin images on cheek part of face with enhanced visualization of clinical characteristics. To test the success of the configuration in detecting acne characteristics, redness value of the resulting image is calculated. The resulting image from subsurface imaging using cross polarized light is shown to have the highest value of redness compared to paralllel polarized light photography and no polarized at all. In addition, to support the detection of these characteristics, image segmentation using K-means clustering algorithms is performed. This algorithm is successfully grouping the resulting acne skin images into five clusters, two of which have the largest redness values so that they are considered to contain images of acne characteristics. After the segmentation process, the area of the two clusters is then calculated using the algorihm to obtain the area of acne on the cheek. This is done to support doctor’s objective assessment regarding the severity of acne and efficacy of acne medication. After the area calculation is successful, a simulation is carried out in which acne images obtained is reduced in size to a certain size which then the area will be calculated. This aims to obtain the resolution of the smallest size for detection and calculation of area of an acne. Acne detection resolution is obtained at 0.34 cm2 with an accuracy of 81%. |
|---|