| Summary: | Iris pigment spot is a discrete pigmentation on the iris surface and can detect eye cancer. There are two types of iris spots, freckles and nevi. While freckles are usually harmless, nevi distort the stromal layer, and therefore, its existence is considered high potential for uveal melanoma, a type of cancer that can cause blindness. The features used to detect the uveal melanoma are size, shape, number of existences, spot of existence and the colour of the pigment spot on the iris surface. In image processing, feature extraction method typically extracts size, shape and colour. However, it is still challenging to produce an accurate extraction result for iris pigment spot. In this study, a threshold intensity value of colour is identified as the pigment spot feature used in the feature extraction process. Furthermore, Spot Filtering Adaptive Thresholding (SFAT) method has been developed to filter between pigment spot and iris surface feature. The proposed method extracted the pigment spot existence with an accuracy rate of 37.1%. This shows that the intensity of the saturation component has the potential to be used in the medical imaging analysis. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
|
|---|