magistrsko delo
Abstract
Možgani so najkompleksnejši organ v človeškem telesu, ki ga kljub ogromnem številu raziskav, še vedno zelo slabo razumemo. Funkcijo možganov običajno raziskujemo skozi različne signale, ki jih generirajo možgani. Eden izmed najpogostejših načinov za merjenje teh signalov je magnetna resonanca. Z analizo funkcijske konektivnosti želimo ugotoviti katere možganske regije se medsebojno odvisne pri proženju nevronov in posledično ugotovimo kako so funkcijsko povezane. Obstoječa literatura nam ponuja številne metrike za izračun funkcijske konektivnosti, a njihova uporaba je nekonsistentna. V okviru naloge smo implementirali, testirali in primerjali razširjene in uveljavljene metrike funkcijske konektivnosti. Metrike smo primerjali po času izvajanja, odpornosti na šum, odpornosti na zamik in pravilnosti zaznavanja vzročnosti. Naši rezultati potrjujejo dejstvo, da so za različne probleme primerne različne metrike, kljub temu pa smo ugotovili, da so najboljše rezultate dosegale metrike Pearsonov koeficient, inverzna kovarianca ter navzkrižna korelacija.
Keywords
funkcijska konektivnost;fMRI;metrike;časovne vrste;magisteriji;
Data
Language: |
Slovenian |
Year of publishing: |
2022 |
Typology: |
2.09 - Master's Thesis |
Organization: |
UL FRI - Faculty of Computer and Information Science |
Publisher: |
[J. Bevc] |
UDC: |
004(043.2) |
COBISS: |
124810499
|
Views: |
827 |
Downloads: |
109 |
Average score: |
0 (0 votes) |
Metadata: |
|
Other data
Secondary language: |
English |
Secondary title: |
Comparison of metrics for functional connectivity |
Secondary abstract: |
The brain is the most complex organ in the human body, which, despite extensive research, is still relatively poorly understood. Brain function is usually investigated by analysing various signals generated by the brain. One of the most common ways to measure these signals is with functional magnetic resonance imaging. By analyzing functional connectivity, we want to find out which brain regions are mutually dependent when firing the neurons and, as a result, find out how they are functionally connected. The existing literature provides us with many metrics for calculating functional connectivity, but their use is inconsistent. As part this work, we implemented, tested and compared widely used and established functional connectivity metrics. The metrics were compared by execution time, noise resistance, lag resistance and correctness of detected causality. Our results confirm the fact that different metrics are suitable for different problems, however, we found that the combination of Pearson's coefficient, inverse covariance and cross-correlation achieved the best results. |
Secondary keywords: |
functional connectivity;fMRI;metrics;time series;computer science;computer and information science;master's degree;Računalništvo;Univerzitetna in visokošolska dela; |
Type (COBISS): |
Master's thesis/paper |
Study programme: |
1000471 |
Embargo end date (OpenAIRE): |
1970-01-01 |
Thesis comment: |
Univ. v Ljubljani, Fak. za računalništvo in informatiko |
Pages: |
48 str. |
ID: |
16587693 |