Background: Digital image analysis has the potential to address issues surrounding traditional histological\r\ntechniques including a lack of objectivity and high variability, through the application of quantitative analysis. A key\r\ninitial step in image analysis is the identification of regions of interest. A widely applied methodology is that of\r\nsegmentation. This paper proposes the application of image analysis techniques to segment skin tissue with\r\nvarying degrees of histopathological damage. The segmentation of human tissue is challenging as a consequence\r\nof the complexity of the tissue structures and inconsistencies in tissue preparation, hence there is a need for a new\r\nrobust method with the capability to handle the additional challenges materialising from histopathological damage.\r\nMethods: A new algorithm has been developed which combines enhanced colour information, created following\r\na transformation to the L*a*b* colourspace, with general image intensity information. A colour normalisation step is\r\nincluded to enhance the algorithm�s robustness to variations in the lighting and staining of the input images. The\r\nresulting optimised image is subjected to thresholding and the segmentation is fine-tuned using a combination of\r\nmorphological processing and object classification rules. The segmentation algorithm was tested on 40 digital\r\nimages of haematoxylin & eosin (H&E) stained skin biopsies. Accuracy, sensitivity and specificity of the algorithmic\r\nprocedure were assessed through the comparison of the proposed methodology against manual methods.\r\nResults: Experimental results show the proposed fully automated methodology segments the epidermis with a\r\nmean specificity of 97.7%, a mean sensitivity of 89.4% and a mean accuracy of 96.5%. When a simple user\r\ninteraction step is included, the specificity increases to 98.0%, the sensitivity to 91.0% and the accuracy to 96.8%.\r\nThe algorithm segments effectively for different severities of tissue damage.\r\nConclusions: Epidermal segmentation is a crucial first step in a range of applications including melanoma\r\ndetection and the assessment of histopathological damage in skin. The proposed methodology is able to segment\r\nthe epidermis with different levels of histological damage. The basic method framework could be applied to\r\nsegmentation of other epithelial tissues.
Loading....