haku: @keyword sparse learning / yhteensä: 1
viite: 1 / 1
« edellinen | seuraava »
Tekijä: | Cao, Yang |
Työn nimi: | Multi-variate analysis of brain images in seasonal affective disorder |
Julkaisutyyppi: | Diplomityö |
Julkaisuvuosi: | 2013 |
Sivut: | 63 s. + liitt. 6 Kieli: eng |
Koulu/Laitos/Osasto: | Perustieteiden korkeakoulu |
Oppiaine: | Informaatiotekniikka (T-61) |
Valvoja: | Oja, Erkki |
Ohjaaja: | Leemput, Koen Van |
OEVS: | Sähköinen arkistokappale on luettavissa Aalto Thesis Databasen kautta.
Ohje Digitaalisten opinnäytteiden lukeminen Aalto-yliopiston Harald Herlin -oppimiskeskuksen suljetussa verkossaOppimiskeskuksen suljetussa verkossa voi lukea sellaisia digitaalisia ja digitoituja opinnäytteitä, joille ei ole saatu julkaisulupaa avoimessa verkossa. Oppimiskeskuksen yhteystiedot ja aukioloajat: https://learningcentre.aalto.fi/fi/harald-herlin-oppimiskeskus/ Opinnäytteitä voi lukea Oppimiskeskuksen asiakaskoneilla, joita löytyy kaikista kerroksista.
Kirjautuminen asiakaskoneille
Opinnäytteen avaaminen
Opinnäytteen lukeminen
Opinnäytteen tulostus
|
Sijainti: | P1 Ark Aalto 7113 | Arkisto |
Avainsanat: | neuroimaging univariate method multivariate method sparse learning SVMs VMB RVM RVoxM SPM seasonal affective disorder |
Tiivistelmä (eng): | In this dissertation, I present different approaches of statistical analysis which can be used in medical research for both structural magnetic imaging and functional imaging. In particular, I study univariate and multivariate methods and experimented using brain imaging data. Methods mainly talked in my thesis are voxel-based morphometry (VBM), support vector machine (SVM), relevance vector machine (RVM), and relevance voxel machine (RVoxM). VBM is a widely used univariate method which is already implanted in the software of MATLAB within the SPM package. SVM and RVM are two existing multi-variate methods which can solve both classification and regression problems regarding to the brain image scans, and the differences between these two methods are such that the first one is easier to handle in practice while the latter offers more sparse results. Another multi-variate analysis method RVoxM takes neighbourhood into consideration when comparing with RVM and therefore offers a more reasonable way to think about the future of neuroimaging, especially in the field of functional imaging. Thus in the first part of my thesis, I introduce the methodology of the univariate method and multivariate methods. Together with methodology, I provide the comparison of each approach. As theory should serve for practice, in the second part of my thesis, I describe the experiments I have done using the mentioned methods with anatomical data and functional data. I focus on seasonal affective disorder (SAD) when facing functional imaging problem, for the situation of SAD in Finland is quite severe. After that I offer a discussion and conclusion. All in all, the univariate and multi-variate methods do excellent jobs in solving neuroimaging problems. In the future, the theory will be more mature and so is the corresponding implementation of engineering. |
ED: | 2013-02-25 |
INSSI tietueen numero: 45846
+ lisää koriin
« edellinen | seuraava »
INSSI