University of Ioannina
Collagen morphometric analysis in liver biopsies of patients with chronic viral hepatitis (HCV or HBV) predicts virological response
Abstract
dc:descriptionExtraction of collagen proportionate area (CPA) in liver biopsy images provides the amount of fibrosis in liver tissue. This is the most distinguishing histological feature in viral hepatitis. Staging is currently based on semi-quantitative scores, such as the Ishak and METAVIR. CPA calculation based upon image analysis techniques has since proven to be more accurate than semi-quantitative scores. Though, lack of standardization and robust methodologies for assessment of computerized image analysis for CPA has proven to be a major limitation and hence CPA has not yet reached daily clinical practice. This current work proposes a fully automated methodology for CPA extraction. It is based on machine learning techniques and is composed by three stages. In particular, background tissue separation and fibrosis detection in regions of liver tissue has been performed with the use ofclustering algorithms. Classification algorithms have also been employed in order to differentiate between liver tissue regions and non-liver tissue regions, such as structural collagen, muscle tissue and blot clots. The non-liver tissue regions have been then excluded from the CPA computation. The methodology was evaluated with the use of 79 liver biopsy images from patients with hepatitis C. The obtained mean absolute CPA error was 1.31% with a concordance correlation coefficient of 0.923.Manual threshold-based and region selection processes, which are widely used in literature, are being avoided with the proposed methodology. Moreover, the CPA calculation time is minimized. In regards to the use of the technique in patients with hepatitis B, we concluded that digital image analysis requires further evaluation.
Degree
thesis:*- Grantor dc:publisher
- University of Ioannina
- Year dc:date
- 2019
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Tsianou, Zoi
- Τσιάνου, Ζωή
Subjects
dc:subject × 26- Ποσοτικοποίηση κολλαγόνου ως αναλογία κολλαγόνου
- Βιοψία ήπατος
- Ηπατική ίνωση
- Ιογενής ηπατίτιδα C
- Ιογενής ηπατίτιδα B
- Κατηγοριοποίηση ηπατικής ίνωσης
- Ψηφιακή ανάλυσης εικόνων
- Μηχανική μάθηση
- Αλγόριθμοι ομαδοποίησης
- Αλγόριθμοι κατηγοριοποίησης
- Collagen proportionate area (CPA)
- Liver biopsy
- Hepatic fibrosis
- Viral hepatitis C
- Viral hepatitis B
- Digital image analysis
- Liver fibrosis staging
- Machine learning
- Clustering algorithms
- Classification algorithms
- Ιατρική και Επιστήμες Υγείας
- Κλινική Ιατρική
- Επιστήμες Υγείας
- Medical and Health Sciences
- Clinical Medicine
- Health Sciences
Rights
dc:rights- Statement dc:rights
-
- BY
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/47322