2026-09-19T17:11:30Zhttps://dspace.mit.edu/server/oai/request

oai:dspace.mit.edu:1721.1/792362022-01-13T07:54:01Zcom_1721.1_7582com_1721.1_7581col_1721.1_131023

Polina Golland. Sweet, Andrew (Andrew Douglas) Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science. Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science 2013-06-17T19:49:40Z 2013-06-17T19:49:40Z 2013 2013 http://hdl.handle.net/1721.1/79236 845314561 Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2013. Cataloged from PDF version of thesis. Includes bibliographical references (p. 51-55). In this thesis, we study approaches for detecting anomalous regions in brain connectivity networks estimated from resting state fMRI. We are motivated by the problem of localizing diseased regions to be resected in pre-surgical epilepsy patients. Our goal is to investigate the potential of these non-invasive connectivity approaches to augment and even replace the clinical gold standard for localization, which requires invasive implantation of electrodes onto the surface of the brain. We focus on adapting an existing method that detects anomalies from a small set of large candidate regions in a population of patients. The main contribution of the work is to develop this method for our application, so that it can efficiently identify anomalies from a large set of small candidate regions in a single epilepsy patient. We find that standard statistical approaches identify regions that overlap reasonably well with electrode recordings of abnormal activity, but are sensitive to manual parameter selection. Our method matches this performance, but has the advantage of automatically determining its corresponding parameters. While localization is not generally accurate enough to consider replacement of invasive electrode implantation, the method discovers potentially diseased regions that may better guide electrode placement. by Andrew Sweet. S.M. x, 55 p. eng Massachusetts Institute of Technology M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission. http://dspace.mit.edu/handle/1721.1/7582 Electrical Engineering and Computer Science. Anomaly detection in brain connectivity structure : an application to epilepsy Thesis application/pdf false Publication unknown unknown <Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="5265f9cb-05a8-4f0c-bbe8-f98e6c6a9e1c"> <Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type> <Language>eng</Language> <Title>Anomaly detection in brain connectivity structure : an application to epilepsy</Title> <PublishedIn> <Publication> </Publication> </PublishedIn> <PublicationDate>2013</PublicationDate> <Authors> <Author> <DisplayName>Sweet, Andrew (Andrew Douglas)</DisplayName> <Affiliation> <OrgUnit> </OrgUnit> </Affiliation> </Author> </Authors> <Editors> </Editors> <Publishers> <Publisher> <DisplayName>Massachusetts Institute of Technology</DisplayName> <OrgUnit /> </Publisher> </Publishers> <License>http://dspace.mit.edu/handle/1721.1/7582</License> <Keyword>Electrical Engineering and Computer Science.</Keyword> <Abstract>In this thesis, we study approaches for detecting anomalous regions in brain connectivity networks estimated from resting state fMRI. We are motivated by the problem of localizing diseased regions to be resected in pre-surgical epilepsy patients. Our goal is to investigate the potential of these non-invasive connectivity approaches to augment and even replace the clinical gold standard for localization, which requires invasive implantation of electrodes onto the surface of the brain. We focus on adapting an existing method that detects anomalies from a small set of large candidate regions in a population of patients. The main contribution of the work is to develop this method for our application, so that it can efficiently identify anomalies from a large set of small candidate regions in a single epilepsy patient. We find that standard statistical approaches identify regions that overlap reasonably well with electrode recordings of abnormal activity, but are sensitive to manual parameter selection. Our method matches this performance, but has the advantage of automatically determining its corresponding parameters. While localization is not generally accurate enough to consider replacement of invasive electrode implantation, the method discovers potentially diseased regions that may better guide electrode placement.</Abstract> <Access xmlns="http://purl.org/coar/access_right" > </Access> </Publication>