Baker’s cyst classification using random forests
Authors:
- Adam Ciszkiewicz,
- Grzegorz Milewski,
- Jacek Lorkowski
Abstract
In this paper, a classification procedure forBaker’s cysts was proposed. The procedure contained twosubprocedures: the image preprocessing (dual thresholding,labeling, feature extraction) and the classification (RandomForests, cross validation). In total, five features were requiredto classify the cysts. These geometric features represented thelocation, the area and the convexity of the cyst. The procedurewas proven effective on a set 436 varied MRI images. The setcontained 68 images with cysts ready for aspiration and wasoversampled with the SMOTE approach. The proposed methodoperates on 2D MRI images. This reduces the time of diagnosisand, with the ever increasing demand for MRI scanners, isjustified economically. The method can be employed in systemsfor autonomous and semi-autonomous Baker’s cyst aspirationor as a standalone package for MRI images annotation.Furthermore, it can be also extended to other fluid-basedmedical conditions in the knee.
- Record ID
- CUT45429514044d48d1b9d9289bd0be6cef
- Publication categories
- ; ;
- Author
- Pages
- 97-100
- Other elements of collation
- rys.; tab.; Bibliografia (na s.) - 100; Bibliografia (liczba pozycji) - 22; Oznaczenie streszczenia - Abstr.
- Book
- Ganzha Maria, Maria Ganzha Maciaszek Leszek, Leszek Maciaszek Paprzycki Marcin Marcin Paprzycki (eds.): Proceedings of the 2018 Federated Conference on Computer Science and Information Systems, September 9–12, 2018, Poznań, Poland, Annals of Computer Science and Information Systems, no. 15, 2018, Warszawa, New York City, Institute of Electrical and Electronics Engineers, Polskie Towarzystwo Informatyczne, Institute of Electrical and Electronics Engineers, ISBN 978-83-949419-5-6 (Web)
- Keywords in English
- biomedical MRI, feature extraction, image classification, image segmentation, medical image processing, random functions
- DOI
- DOI:10.15439/2018F89 Opening in a new tab
- URL
- https://annals-csis.org/Volume_15/index.html Opening in a new tab
- Language
- eng (en) English
- License
- Score (nominal)
- 20
- Additional fields
- Indeksowana w: Web of Science, Scopus
- Uniform Resource Identifier
- https://cris.pk.edu.pl/info/article/CUT45429514044d48d1b9d9289bd0be6cef/
- URN
urn:pkr-prod:CUT45429514044d48d1b9d9289bd0be6cef
* presented citation count is obtained through Internet information analysis, and it is close to the number calculated by the Publish or PerishOpening in a new tab system.