National and Kapodistrian University of Athens
Εφαρμογή νέων μεθόδων κατατμήσεως ψηφιακής εικόνας στην μελέτη του καρκινώματος τους ουροδόχου κύστεως
Abstract
dc:descriptionINTRODUCTION: To develop a method for non‐supervised thresholding of transitional cell carcinoma nuclei of routinely hematoxylin‐eosin (H‐E) stained histological sections, as well as a method for automatic selection of immunohistochemically positive areas. STUDY DESIGN: For greyscale, RGB, HSL and L*a*b* thresholding we used an extension of a clustering method, based on a between‐class / within‐class criterion proposed by Otsu, applying an optimal grey‐level thresholding to each of the distributions of the R, G, B, or H and S or L* color domains. The algorithms were tested on clinical material consisted of 20 H‐E stained sections of bladder carcinomas. The results from the 4 thresholding techniques were evaluated in comparison with the corresponding results of manually selected nuclear areas. For automatic evaluation of immunohistochemical images, a clustering algorithm in L*a*b* color space was developed. RESULTS: The results from the 4 thresholding techniques were evaluated in comparison with the corresponding results of manually selected nuclear areas. Thus, every image was compared pixel to pixel with its matching reference image. Greyscale automatic thresholding presented an unacceptably low pixel specificity, which complicated the further nuclear segmentation. Non‐supervised thresholding in RGB or HSL, as well as semi‐manual thresholding in L*a*b* colorspace demonstrated significantly better accuracy and high values of pixel specificity and sensitivity, which permitted errors of only 4.27‐5.83% in the subsequent mean area estimation of the transitional cell carcinoma nuclei. CONCLUSION: The non‐supervised multispectral thresholding in RGB or HSL colorspace as presented here extends single grey‐level thresholding techniques to multilevel thresholding. This seems to be a very effective, relatively simple and fast alternate to the widely used automatic greyscale or manual color thresholding for the segmentation of nuclei in routine histological sections.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2013
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Pavlopoulos, Petros
- Παυλόπουλος, Πέτρος
Subjects
dc:subject × 18- Ανάλυση εικόνας
- Ουροθηλιακό καρκίνωμα ουροδόχου κύστης
- Κατάτμηση πυρήνων
- Συστάδες
- Εφαρμογή ουδού
- Image analysis
- Bladder carcinoma
- Nuclear segmentation
- Clustering
- Thresholding
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Ιατρική και Επιστήμες Υγείας
- Βασική Ιατρική
- Natural Sciences
- Computer and Information Sciences
- Medical and Health Sciences
- Basic Medicine
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/29838