| Title |
Unsupervised machine learning methods for neonatal EEG: application and economic benefit insights |
| Translation of Title |
Neprižiūrimo mašininio mokymosi metodai naujagimių EEG signalams analizuoti: taikymas ir ekonominės naudos įžvalgos. |
| Authors |
Abraškevičiūtė, Vaida |
| Full Text |
|
| Pages |
59 |
| Keywords [eng] |
EEG ; neonatal ; hypoxic-ischaemic encephalopathy ; unsupervised machine learning |
| Abstract [eng] |
This paper examines the application of unsupervised machine learning algorithms for EEG analysis, as well as their economic considerations. It focuses on EEGs of neonates diagnosed with hypoxic-ischemic encephalopathy (HIE). A total of 169 one-hour epochs were used for unsupervised clustering. The study aims to apply unsupervised machine learning algorithms to cluster the EEGs into grades based on the severity of HIE by analyzing the EEG background. Various types of preprocessing, feature selection methods, and algorithms are considered, and their performance is evaluated. Moreover, this paper examines the economic benefits that these innovations can bring. |
| Dissertation Institution |
Kauno technologijos universitetas. |
| Type |
Master thesis |
| Language |
English |
| Publication date |
2025 |
