Institutional Repository of Vilnius University
Tekstūros auglių klasifikavime neuroniniais tinklais /
Abstract
dc:descriptionThe work examines the value of textural information for the task of brain tumor classification using neural networks. In order to extract richer features from MRI images, edge detection and statistical texture extraction methods are applied to the images across three different area slices (full image, tumor segment, expanded tumor segment), which are then used to train convolutional neural networks. The addition of edge detection based and statistical textural information is found to greatly improve the accuracy of convolutional neural networks when classifying brain tumor segments, statistical textural information is found to improve the accuracy of full image classification. The model trained on data consisting of expanded tumor segments achieved the greatest accuracy, yet the model was not improved by applying the textural methods used in the work – further research of expanded tumor segment texture information is possible.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Samoilovas, Ignas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD146242340&prefLang=en_US