Institutional Repository of Vilnius University
Dirbtinių neuroninių tinklų taikymas plaučių uždegimo atpažinimui /
Abstract
dc:descriptionThis work explores the use of artificial neural networks to detect pneumonia from X‑ray images. It reviews existing solutions for using neural networks to classify pneumonia, which usually only classifies whether a person has pneumonia or not, but does not determine the type. In this work, by modifying existing neural networks for two‑class classification, models are developed that can identify four classes: healthy lung, bacterial pneumonia, viral pneumonia and COVID‑ 19 pneumonia. These modified networks are also shown to be able to accurately classify different types of pneumonia compared to their original two‑class classification models. It also identifies the problem that there are no ready‑made datasets for multiclass classification of pneumonia, and that by combining several different publicly available datasets, the number of labeled images of bacterial and viral types is limited.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2023
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Kemešis, Domantas,
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:ELABAETD192846617&prefLang=en_US