Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
Automated Nuclei Analysis from Digital Histopathology
Ist Teil von
2023 International Conference on Intelligent Systems, Advanced Computing and Communication (ISACC), 2023, p.1-6
Ort / Verlag
IEEE
Erscheinungsjahr
2023
Quelle
IEEE Electronic Library Online
Beschreibungen/Notizen
In Histopathology analysis, an abnormal nuclear shape can be a strong parameter to detect malignancy. Similarly, by visualizing the growing amount of nuclei implies disease status such as grading of cancer. In clinical diagnosis there is a link between texture of the nucleus and disease current status. In digital pathology the most crucial & important step for computer-aided diagnostics system is automated cell nuclei segmentation. Regular studies are going on digital pathology for automatic analyzing the nuclei cell images. However, providing accurate cell nucleus segmentation rate through automation is still poor. Scientists are rigorously working on the development of computerized segmentation algorithm to provide faster diagnosis. This paper proposed a novel thresholding based unsupervised technique which is applicable to detect nuclei contours. Proposed approach is applied to seven multi organ histological tissue images. Threshold is calculated by taking into consideration the first nonzero gray level pixel value followed by finding the pixel connectivity with 8 neighborhood concept. Furthermore, the various nuclei feature including mean intensity, diameter, perimeter, centroid is computed to get better insight to cell functionality providing a support in early diagnosis to diseases. When compare with traditional adaptive thresholding this approach achieved better performance rate by 48.31 %.