[dipl] Perustieteiden korkeakoulu / SCI

Permanent URI for this collectionhttps://aaltodoc.aalto.fi/handle/123456789/21

Browse

Browsing [dipl] Perustieteiden korkeakoulu / SCI by Department "Lääketieteellisen tekniikan ja laskennallisen tieteen laitos"

Filter results by typing the first few letters
Now showing 1 - 20 of 31
  • Adapting a Gaze Tracking System to Mobile Environments
    (2012) Meriläinen, Arto
     Master's thesis
    Gaze tracking has traditionally been performed in controlled laboratory environments. The experiment setups have commonly been limited to study only computer-human interaction. Recently, the need to perform experiments in natural environments has emerged in different areas of science. This thesis overviews the structure of a mobile gaze tracking system, that was developed in the Ganzheit project. The goal of the gaze tracking system is to offer an open alternative for commercial systems. The developed gaze tracker utilizes model-based gaze tracking. The approach is accurate and robust against movements of the head. In order to utilize model-based gaze tracking, it is vital to identify the pupil and corneal reflections from an eye image. The constructive part of the thesis focuses on developing a method for identifying the features from the eye image. The developed method is tested with experimental data. The results show that the method for finding the pupil is accurate and robust against differences in facial features and changing lighting conditions. The goodness of recognizing the corneal reflections varies between test subjects. The mobile gaze tracking system is experimented in an ordinary office room. The results indicate that the developed method works adequately.
  • Automatic seizure detection using a two-dimensional EEG feature space
    (2011) Tanner, Antti
     Master's thesis
    Epileptinen kohtaus on neurologinen häiriötila, joka ilmenee aivojen epänormaalina sähköisenä toimintana. Joihinkin kohtauksiin liittyy ulkoisia merkkejä, kuten lihaskouristuksia. Kohtauksia, joihin ei liity selkeitä ulkoisia merkkejä, kutsutaan ei-konvulsiiviksi. Ne voidaan tunnistaa vain seuraamalla aivojen sähköistä toimintaa. Ei-konvulsiivisten kohtauksien on osoitettu olevan erityisen yleisiä tehohoitopotilailla - myös sellaisilla potilailla, joilla ei ole aiemmin ollut kohtauksia. Epileptinen kohtaus on pikaista interventiota vaativa vakava tila. Aivosähkökäyrällä (elektroenkefalografia, EEG) voidaan tutkia aivojen sähköistä toimintaa. Datan läpikäynti käsin on aikaavievää, joten tehohoitoon sopivalle, automaattiselle ja reaaliaikaiselle analyysimenetelmälle on suuri tarve. Tässä diplomityössä esitellään kolme menetelmää, jotka soveltuvat signaalipiirteiden evoluution seuraamiseen. Kultakin EEG-kanavalta määritetään kaksi piirrettä: hetkellinen taajuus ja signaalin teho. Ensimmäinen menetelmä mittaa piirreavaruuteen muodostuvan polun pituutta aikatasossa. Toinen menetelmä vertaa kutakin piirreavaruudessa otettua askelta edellisiin askeliin. Kolmannessa menetelmässä määritetään dynaamisesti edellisistä piirrevektoreista konveksi kuori ja tutkitaan kuoren ulkopuolelle osuvia piirrevektoreita. Kolmas menetelmä osoittautui tutkimuksessa parhaaksi. Menetelmällä pystyttiin tunnistamaan 11 tietokannan 19:sta kohtauksista kärsineestä potilaasta. Tietokannassa on EEG-mittauksia 179 tehohoitopotilaalta. Suurin osa vääristä detektioista johtui EEG:ssä näkyvästä lihastoiminnasta, artefaktoista tai alkeellisesta tunnistuslogiikasta. Menetelmän todellista suorituskykyä on liian aikaista arvioida. Menetelmää pitää täydentää EEG-piikit sekä artefaktat luotettavasti tunnistavilla algoritmeilla.
  • Bayesian Classification of fMRI Patterns for Natural Audiovisual Stimuli Using Sparsity Promoting Laplace Priors
    (2012) Koistinen, Olli-Pekka
     Master's thesis
    Bayesian linear binary classification models with sparsity promoting Laplace priors were applied to discriminate fMRI patterns related to natural auditory and audiovisual speech and music stimuli. The region of interest comprised the auditory cortex and some surrounding regions related to auditory processing. Truly sparse posterior mean solutions for the classifier weights were obtained by implementing an automatic relevance determination method using expectation propagation (ARDEP). In ARDEP, the Laplace prior was decomposed into a Gaussian scale mixture, and these scales were optimised by maximising their marginal posterior density. ARDEP was also compared to two other methods, which integrated approximately over the original Laplace prior: LAEP approximated the posterior as well by expectation propagation, whereas MCMC used a Markov chain Monte Carlo simulation method implemented by Gibbs sampling. The resulting brain maps were consistent with previous studies for simpler stimuli and suggested that the proposed model is also able to reveal additional information about activation patterns related to natural audiovisual stimuli. The predictive performance of the model was significantly above chance level for all approximate inference methods. Regardless of intensive pruning of features, ARDEP was able to describe all of the most discriminative brain regions obtained by LAEP and MCMC. However, ARDEP lost the more specific shape of the regions by representing them as one or more smaller spots, removing also some relevant features.
  • Bayesian Regression Analysis of Sickness Absence
    (2011) Kahilakoski, Olli-Pekka
     Master's thesis
    Gaussiset prosessit ovat epälineaarisia regressiomalleja, joilla voidaan mallintaa paikallisia muutoksia vastepinnan rakenteessa. Sairauspoissaoloihin yhteydessä olevia yksilötekijöitä on aiemmin tutkittu yleistetyillä lineaarimalleilla. Vertaamme tässä työssä gaussisia prosesseja yleistettyihin lineaarimalleihin bayesilaisilla menetelmillä ja havaitsemme, että gaussiset prosessit ennustavat yleistettyjä lineaarimalleja paremmin sairauspoissaoloja terveyskyselyn avulla. Teemme myös muuttujanvalinnan gaussisille prosesseille bayesilaisella monivertailumenetelmällä ja havaitsemme, että masennuksella ja kivun aiheuttamalla työhaitalla on yhteys sairauspoissaoloihin. Tulokset ovat linjassa aiempien tutkimusten kanssa. Lisäksi havaitsemme masennuksella ja sairauspoissaoloilla mahdollisen epälineaarisen, saturoituvan yhteyden.
  • Brain state dynamics in transcranial magnetic stimulation - A combined TMS-EEG study
    (2013) Mutanen, Tuomas
     School of Science | Master's thesis
    Transcranial magnetic stimulation (TMS) and electroencephalography (EEG) have been successfully combined to study the connectivity and reactivity of the brain. However, it is not yet well understood how TMS modulates the ongoing brain activity largely because the present methods used to analyze TMS-EEG signals usually describe the average response to TMS rather than the immediate effects. The purpose of this Thesis is to improve our understanding on the dynamics of TMS by analyzing EEG signals; How does TMS affect the state of the brain, and on the other hand, how does the state of the brain change the effects of TMS? Deeper understanding on this subject is vital when seeking for more elaborate and effective stimulation sequences and methods. In this Thesis, we introduce two quantitative tools called mean state shift (MSS) and state variance (SV) and show that they are able to quantify the transient effects of TMS on the electric brain state. Furthermore, by performing measurements where the state of the brain was modulated before the actual test TMS pulse we show that this state modulation affects post-TMS EEG. Furthermore, the group level results imply that the TMS-elicited changes in MSS and SV are sensitive to the pre-TMS state modulation.
  • Classifying the brain responses to hand action video clips with fMRI: application to complex regional pain syndrome diagnostics
    (2013) Saari, Jukka
     School of Science | Master's thesis
    Functional magnetic resonance imaging (fMRI) is one of the most powerful tools in modern neuroscience. In multi-variate pattern analysis (MVPA) of fMRI data, statistical machine learning classifiers are used to decode the content of the presented stimuli from the spatial brain activity patterns. In this thesis, healthy controls and patients suffering from complex regional pain syndrome (CRPS) affecting one hand were shown hand action video clips during fMRI scanning. Observing, or even imagining, hand actions causes unpleasantness and increases the pain levels in CRPS patients. The aim was to study, with MVPA, how abnormal processing of the observed hand actions was reflected in the fMRI data. Subject-wise video clip category and group-level patient vs. control classifications were performed. As a result, classification of the video clip category (either a hand making a fist or squeezing an object), with up to 70 % accuracy on average, was possible in the primary and the secondary somatosensory cortices and other areas associated with action observation. The group-level analyses revealed that the patients were possible to distinguish from the controls, with 80-100 % accuracy, in brain areas associated with pain (e.g. anterior cingulate cortex) and action observation based on the brain responses elicited by observing the hand action video clips. Also, patients' individual video clip classification accuracies in the primary and the secondary somatosensory cortices correlated negatively with the self-reported pain levels and the level of upper limb disability and positively with the patients' age. These results show that MVPA is a valuable method in subject-wise classification analyses as well as in group-level patient vs. control classification.
  • Comparison between patient spirometry and ventilator spirometry
    (2011) Jenu, Saana
     School of Science | Master's thesis
  • Kaasujenvaihdon mittausten verifikaatiolaitteiston suunnittelu
    (2011) Tommila, Rurik
     School of Science | Master's thesis
    GE Healthcare Finland site in Vallila produces different kinds of gas modules, one of which is E-COVX designed to measure patient's gas exchange. The module provides information which may be vital to the patient, and therefore they must be tested before sending to customer. This testing has traditionally been taken care of with a hardware based on Deltatrac II metabolic monitor, heralded as a golden standard when measuring gas exchange. However, the service of the monitor will be shut down, and therefore new solutions for testing the modules must be sought. This thesis presents a solution to replace the old hardware. The goal was to achieve at least the same accuracy as the old hardware and improve its usability. Based on these goals, the chance to acquire a commercial solution is studied. When no such solution is found, we design and execute a whole new system for testing the modules. Based on the preliminary verification measurements concluded on the new solution, it can be said to achieve almost the same accuracy as the Deltatrac and clearly better accuracy than the modules tested. There is still room for improvement: The measurement can be fully automatized and more accurate components can be acquired in order to improve overall accuracy.
  • ECG Parameters in Short-Term Prediction of Ventricular Arrhythmias
    (2012) Kemppainen, Reko
     Master's thesis
    Malignant spontaneous ventricular arrhythmias, such as ventricular tachycardia (VT) and ventricular fibrillation (VF), are the most common trigger of sudden cardiac death (SCD) in and out of hospital. For a hospitalized patient, occurrence of such arrhythmia is a struggle of life and death where every second of oxygen deprivation, resulting from reduced blood flow, decreases chances of survival. Despite recent advances in resuscitation strategies, survival rates in in-hospital cardiac arrests remain unacceptably low. Main factors contributing to the poor prognosis are lack of patient monitoring and delay in the initiation of resuscitation. Thus, in order to increase the likelihood of successful resuscitation, or prevent the arrhythmia from happening in the first place, continuous and quantitative risk of arrhythmia assessment is required. Currently, however, cardiac monitoring is utilized to detect the onset of life threatening cardiac episodes only. Thus, development of risk indices and the study of precursors of lethal arrhythmias have great clinical value and will lead to better cardiac monitoring. In this thesis, changes in ECG signal preceding lethal cardiac arrhythmias are studied both in different patient groups and in individual patients. Furthermore, an algorithm predicting imminent ventricular tachyarrhythmias is presented. Current knowledge of underlying mechanisms of onset of ventricular arrhythmias is used to assess the risk of arrhythmia continuously during cardiac monitoring of a patient. Our approach is novel and similar assessment of such algorithm has not been published previously. A review of existing methods and applications for risk assessment of SCD with discussion of future trends and possibilities is also given.
  • Rekoveriinin vaikutus hiiren sauvasolun valovasteisiin
    (2012) Pitkänen, Marja
     School of Science | Master's thesis
    Fotonin absorboituminen verkkokalvon näköaistinsolun näköpigmenttimolekyyliin rodopsiiniin käynnistää solussa fototransduktion. Tämä biokemiallinen tapahtumasarja muuttaa tiedon valon saapumisesta sähköiseksi hermosignaaliksi. Rekoveriini on näköaistinsoluissa sijaitseva kalsiumsensoriproteiini, joka vaikuttaa fototransduktion vahvistukseen säätelemällä rodopsiinin inaktivaatiota. Tässä diplomityössä tutkittiin rekoveriinin vaikutusta sauvasolun sähköisiin valovasteisiin. Menetelmänä käytettiin eristetyn verkkokalvon elektroretinografiaa (ERG), jossa rekisteröidään näköaistinsolujen valovasteiden aikaansaamia muutoksia verkkokalvon yli mitattavassa jännitteessä. Mittaukset tehtiin C57BL/6-kannan hiirten verkkokalvoilla, jotka olivat poistogeenisiä rekoveriinin suhteen. Vertailuksi samat mittaukset tehtiin myös tavallisten C57BL/6-kannan hiirten verkkokalvoilla. Valovasteista määritettiin useita parametreja, jotka kuvaavat mm. sauvasolun herkkyyttä, fototransduktion aktivaatiovaiheen vahvistusta ja inaktivaatioreaktioiden kinetiikkaa. Tavallisten ja poistogeenisten hiirten parametreja vertailemalla saatiin tietoa rekoveriinin toiminnasta fototransduktion säätelyssä. Mittaustulokset olivat pääosin yhteneväisiä aiempien muilla menetelmillä tehtyjen tutkimusten kanssa: rekoveriinilla oli pieni kasvattava vaikutus sauvasolun herkkyyteen sekä hidastava vaikutus kinetiikkaan. Uutena tuloksena rekoveriinin havaittiin mm. muuttavan valovasteen paluun määräävää aikavakiota. Lisäksi havaittiin, että rekoveriini ei välttämättä ole ainut kalsiumin säätelemä proteiini, joka vaikuttaa fototransduktiossa tapahtuvaan valoadaptaatioon. Näiden tulosten perusteella vaikuttaa siltä, että fototransduktioon saattaa vaikuttaa jokin tuntematon kalsiumpitoisuudesta riippuva säätelymekanismi.
  • Fabrication and Characterization of Thermophotonic Devices
    (2011) Olsson, Anders
     Master's thesis
    Tässä diplomityössä tutkitaan termofotonisille laitteille tarkoitettujen LED-rakenteiden valmistusta ja karakterisointia. Termofotonisen laitteen toiminta perustuu elektroluminesenssijäähdytykseen, eli LEDin kykyyn emittoida fotoneja, jotka ovat saanet osan energiastaan hilalämmöstä. Ilmiö saattaa mahdollistaa LEDin jäähtymisen sen emittoidessa valoa, jos häviöt ovat riittävän pieniä. Työn teoreettisessa osassa käsitellään lyhyesti LEDien ja metallikontaktien perusominaisuuksia. Metallikontakteja käytetään injektoimaan virtaa ja heijastamaan valoa rakenteissa. Valmistusprosessi alkaa indiumfosfidi (InP) puolijohdesubstraatista, jonka pinnalle kasvatetaan useita puolijohdekerroksia, ml. aktiivinen InGaAs-kerros, tavoitteena saada aikaan valoa emittoiva p-n-liitos. Sähköiset kontaktit valmistetaan höyrystämällä metallia puolijohderakenteiden pinnalle. Valmistetut rakenteet karakterisoidaan fotoluminesenssimittauksilla (PL), virtajännitemittauksilla (I-V) ja atomivoimamikroskoopilla (AFM). Näytteitä valmistettiin työn aikana 63 kpl, joista 22 oli valmiita InP/InGaAs p-n-liitos LED-rakenteita joiden PL-huiput sijoittuivat välille 1630 - 1690 nm.
  • Functional localization and investigation of cortical speech areas by navigated TMS and EEG
    (2013) Mäkelä, Niko
     School of Science | Master's thesis
    Transcranial magnetic stimulation (TMS) combined with simultaneous electroencephalography (EEG) is a noninvasive tool for measuring cortical excitability and connectivity. TMS-evoked large muscle artifacts may mask the TMS-evoked shortlatency brain activity, especially when applied on lateral areas. Classical speechrelated cortical areas are located laterally, which limits the usability of TMS-EEG on studying the cortical speech network. The TMS-evoked artifacts can be reduced both on-line and off-line. A carefully performed measurement reduces the level of artifacts significantly. Remaining artifacts can be removed, suppressed, or separated to some extent, from the brain signals with computational methods. In this Thesis, the applicability of TMS-EEG on studying the lateral speech areas was evaluated with a combination of on-line and off-line methods. The largest muscle artifacts occur when TMS is applied on anterior lateral areas. In this study, a posterior speech-related area was located functionally with a novel TMS speech mapping method and used as the stimulation target in TMS-EEG. The effects of low stimulation intensities on the contaminating artifacts were evaluated against the conventional intensities. TMS-evoked EEG was analyzed in the signal, sensor, and source spaces. The artifactual components were separated from the neural components with independent component analysis (ICA). The effects of different preprocessing approaches on the performance of ICA were studied. The results showed that both the amplitude and duration of the muscle artifacts decreased fast when the stimulation intensity was lowered. However, the 20 first milliseconds still remain a challenging task in TMS-EEG of lateral areas. In addition, the results put the use of conventional trial-averaged data in ICA in serious doubt.
  • Gaussian filtering and smoothing based parameter estimation in nonlinear models for sequential data
    (2012) Väänänen, Ville Juhana
     Perustieteiden korkeakoulu | Master's thesis
    State space modeling is a widely used statistical approach for sequential data. The resulting models can be considered to contain two interconnected estimation problems: that of the dynamic states and that of the static parameters. The difficulty of these problems depends critically on the linearity of the model, with respect to the states, the parameters or both. In this thesis we show how to obtain maximum likelihood and maximum a posteriori estimates for the static parameters. Two methods are considered: gradient based nonlinear optimization of the marginal log-likelihood and expectation maximization. The former requires the filtering distributions and the latter both the filtering and the smoothing distributions. When closed form solutions to these distributions are unavailable, we apply efficient Gaussian filtering based methods to obtain approximations. The resulting parameter estimation algorithms are demonstrated by a linear target-tracking model with simulated data and a nonlinear stochastic resonator model with photoplethysmograph data.
  • Hilbert Space Methods in Infinite-Dimensional Kalman Filtering
    (2012) Solin, Arno
     Master's thesis
    Many physical and biological processes include both spatial and temporal features. Spatio-temporal modeling under the machine learning paradigm of Gaussian process (GP) regression has demonstrated prominent results. However, the appealing Bayesian treatment by GP regression is often difficult in practical problems due to computational complexity. In this thesis, methods for writing spatio-temporal Gaussian process regression as infinite-dimensional Kalman filtering and Rauch - Tung - Striebel smoothing problems are presented. These scale linearly with respect to the number of time steps as opposed to the cubic scaling of the direct GP solution. Spatio-temporal covariance functions are formulated as infinite-dimensional stochastic differential equations. Furthermore, it is presented how infinite-dimensional models can be combined with a finite number of observations to an approximative solution. For this, a truncated eigenfunction expansion of the Laplace operator is formed in various domains, of which the n-dimensional hypercube and hypersphere are explicitly written out. The approach in this thesis is primarily application-driven, and therefore three real-world case studies are presented as proof of concept. The feasibility of infinite-dimensional Kalman filtering is demonstrated by forming a spatio-temporal resonator model which is applied to temperature data in two spatial dimensions, and a novel way of modeling the space{time structure of physiological noise in functional brain imaging data is considered in both two and three spatial dimensions.
  • Kuuloaivokuoren tehtäväsidonnaisten aktivaatioiden ajallinen käyttäytyminen: EEG-tutkimus
    (2012) Talja, Suvi
     Perustieteiden korkeakoulu | Master's thesis
    Rinne at al. (2009) reported that human auditory cortex activations measured with fMRI are strongly dependent on the auditory task. Although fMRI has a high spatial resolution its temporal accuracy is not sufficient to examine the activation order of different areas of the auditory cortex during active listening tasks. The present study tested whether these task-dependent activations can be investigated using source estimation of scalp-recorded auditory evoked potentials. Subjects (17) performed pitch discrimination, pitch memory (three difficulty levels in both tasks) or visual tasks during the presentation of sounds. The auditory evoked potentials were recorded using 136 scalp electrodes. Sources of the evoked potentials were modeled using cortically constrained and depth- and orientation-weighted minimum norm estimation. The evoked potentials were modulated by task at 200-700 ms from sound onset. Source estimation revealed stronger activation (350-700 ms) in the anterior auditory cortex of the left hemisphere during pitch discrimination than during visual task with the same sounds. Pitch memory task, in turn, was associated enhanced activation (vs. visual task) in the bilateral inferior parietal lobule (500-650 ms) and decreased activation (vs. visual and discrimination tasks) in the left anterior superior temporal gyrus (200-300 ms). These task-dependent activations are very similar to those reported in the previous fMRI study. Therefore, the results of the thesis indicate that evoked potentials can be used to obtain temporal information on the task-dependent activations of human auditory cortex.
  • Laitteisto ERG-signaalin samanaikaiseen rekisteröintiin näköaistinsolukerroksesta sekä verkkokalvon yli
    (2011) Turunen, Teemu
     Master's thesis
    Valon absorptio verkkokalvon näköaistinsoluissa saa aikaan monivaiheisen biokemiallisten reaktioiden sarjan, jossa valon sisältämä tieto muunnetaan solutason sähköisiksi signaaleiksi. Näiden signaalien aiheuttamia muutoksia verkkokalvon soluvälitilan jänniteprofiilissa voidaan tutkia elektroretinogrammi -tekniikalla (ERG). Näköaistinsolujen ERG-signaalia voidaan rekisteröidä eristetystä verkkokalvosta kahdella tavalla: Erottamalla näköaistinsolujen vaste farmakologisesti ja rekisteröimällä potentiaalieroa koko verkkokalvon yli (TERG) tai rekisteröimällä potentiaalieroa mikroelektrodeilla vain tutkittavasta solukerroksesta (LERG). Diplomityössä suunniteltiin ja rakennettiin laitteisto, jolla voidaan rekisteröidä samanaikaisesti LERG- ja TERG-vasteita. Laitteistoa testattiin koeohjelmalla, jossa verrattiin ulkojäsenkerroksesta rekisteröityjä LERG-vasteita ja näköaistinsolujen TERG-vasteita. Tulosten pohjalta todettiin, että LERG-tekniikka voidaan käyttää näköaistisolujen toiminnan kvantitatiiviseen tutkimukseen. LERG- ja TERG-vasteita vertailtaessa huomattiin, etteivät vasteiden nousunopeus, katkaisukinetiikka, nousuaika huippuarvoon ja suhteelliset amplitudit poikenneet toisistaan merkittävästi. Suurin eroavaisuus oli voimakkailla valostimuluksilla TERG-vasteisiin muodostuva nopea transientti aalto. Tässä työssä tehdyt kokeet vahvistavat, että kyseinen komponentti syntyy näköaistinsolujen sisäjäsenkerroksessa.
  • MEG Coherence Estimates - Sensitivity Analysis and Clinical Application
    (2014) Luoma, Jarkko
     Perustieteiden korkeakoulu | Master's thesis
    The role of synchronization of neuronal oscillatory activity across distinct brain regions has been recently deemed increasingly important for the proper functioning of the brain, and changes in synchronization have been reported in several diseases of the nervous system, including Parkinson's disease (PD). Magnetoencephalography (MEG) enables non-invasive measurement of neuronal activity with adequate temporal resolution to quantify synchronization-related phenomena. One measure of synchronization between two signals is coherence. By first estimating the neuronal sources that generate the MEG signals, we can compute coherence either between the estimated source signals (cortico-cortical coherence) or between the source signals and a signal recorded from a muscle (cortico-muscular coherence). Quantifying synchronization and finding brain networks responsible for motor control can help to understand the underlying mechanisms of PD. However, estimating coherence from noisy MEG data is difficult and may lead to errorenous conclusions. The purpose of this Thesis was to examine, using simulations, the effectiveness of MEG coherence analysis at the source level, especially related to the network responsible for movement control. The source activities were estimated from the MEG data using beamforming. One focus was the likelihood of false positives, and understanding what creates such artifactual coherence. In addition to typical coherence analysis, I used a recent method known as imaginary coherence, which should reduce the false coherence. Finally, I applied these methods to a MEG data set recorded from 14 PD patients during a motor task. These patients had deep brain stimulators, which alleviate the symptoms of the PD. The motivation here was to assert whether the deep brain stimulator causes systematic changes in coherence. The simulations showed that coherence in MEG at the source level was prone to artifacts which passed statistical testing. The results depended critically on the signal-to-noise ratio of the sources, with lower signal-to-noise ratio creating more false coherence in the results. Moreover, some of these artifacts were localized systematically with respect to the coherent sources, which could be explained by the similarity of MEG sensor-level responses to neuronal sources at specific locations. For 10 out of the 14 PD patients analyzed, the highest cortico-muscular coherence maximum was at the motor cortex in a physiologically viable location. In addition, coherence analysis indicated various other coherent brain regions, however, the reliability of these results was uncertain. I did not observe any systematic effect of deep brain stimulation on coherence. Finally, I suggested various possibilities for future research to alleviate the problem of false coherence.
  • Modular structure of functional brain networks during movie viewing and rest
    (2014) Kujala, Rainer
     Perustieteiden korkeakoulu | Master's thesis
    Network science has successfully shed light on the large-scale structure of various complex systems, such as social networks and the Internet. One area of study where the concepts and tools of network science are increasingly applied, is the analysis of brain imaging data, where the goal is to better understand the structure and function of the human brain. fMRI is a brain imaging modality that produces data that is more and more commonly approached from the network perspective. fMRI data itself consists of time series of activations of 3D voxels. Because of this, the data have to be transformed to a meaningful network representation, which involves a number of challenges. One of the challenges is to decide what the nodes of a functional brain network should represent - e.g. voxels, data-driven aggregations of voxels, or anatomical brain areas. There is probably no perfect solution to this problem. In this work, we approach the problem by analyzing functional brain networks both at the level of nodes corresponding to individual voxels and at the level of network modules, which are obtained using data-driven methods for network partitioning. As a test bed for our approach, we use data from fMRI scans of 13 subjects in two experimental conditions, where the subjects are either viewing a movie or resting. In this work we will focus on highlighting how the different conditions are rejected on the differences in the module-level network structure. Based on our analysis, we conclude that data-driven network partitioning can greatly help in understanding the network structure. Especially, we find that the network coarse-graining approach developed in this Thesis is useful in unveiling the overall topology of fine-grained functional brain networks.