{"id":{"repo_id":"birmingham","oai_identifier":"oai:etheses.bham.ac.uk:11672"},"canonical_url":"https://search.ndltd.org/etd/birmingham/oai:etheses.bham.ac.uk:11672","repository":{"repo_id":"birmingham","name":"University of Birmingham","base_url":"https://etheses.bham.ac.uk/cgi/oai2"},"display":{"title":"Frequency domain high density diffuse optical tomography for functional brain imaging","abstract":"Measurements of dynamic near-infrared (NIR) light attenuation across the human head together with model-based image reconstruction algorithms allow the recovery of three-dimensional spatial brain activation maps. Previous studies using high-density diffuse optical tomography (HD-DOT) systems have reported improved image quality over sparse arrays. Modulated NIR light, known as Frequency Domain (FD) NIR, enables measurements of phase shift along with amplitude attenuation. It is hypothesised that the utilization of these two sets of complementary data (phase and amplitude) for brain activity detection will result in an improvement in reconstructed image quality within HD-DOT. However, parameter recovery in DOT is a computationally expensive algorithm, especially when FD-HD measurements are required over a large and complex volume, as in the case of brain functional imaging. Therefore, computational tools for the light propagation modelling, known as the forward model, and the parameter recovery, known as the inverse problem, have been developed, in order to enable FD-HD-DOT. The forward model, within a diffusion approximation-based finite-element modelling framework, is accelerated by employing parallelization. A 10-fold speed increase when GPU architectures are available is achieved while maintaining high accuracy. For a very high-resolution finite-element model of the adult human head with ∼600,000 nodes, light propagation can be calculated at ∼0.25s per excitation source. Additionally, a framework for the sparse formulation of the inverse model, incorporating parallel computing, is proposed, achieving a 10-fold speed increase and a 100-fold memory efficiency, whilst maintaining reconstruction quality. Finally, to evaluate image reconstruction with and without the additional phase information, point spread functions have been simulated across a whole-scalp field of view in 24 subject-specific anatomical models using an experimentally derived noise model. The addition of phase information has shown to improve the image quality by reducing localization error by up to 59%, effective resolution by up to 21%, and depth penetration up to 5mm, as compared to using the intensity attenuation measurements alone. In addition, experimental data collected during a retinotopic experiment reveal that the phase data contains unique information about brain activity and enables images to be resolved for deeper brain regions.","abstract_html":"Measurements of dynamic near-infrared (NIR) light attenuation across the human head together with model-based image reconstruction algorithms allow the recovery of three-dimensional spatial brain activation maps. Previous studies using high-density diffuse optical tomography (HD-DOT) systems have reported improved image quality over sparse arrays. Modulated NIR light, known as Frequency Domain (FD) NIR, enables measurements of phase shift along with amplitude attenuation. It is hypothesised that the utilization of these two sets of complementary data (phase and amplitude) for brain activity detection will result in an improvement in reconstructed image quality within HD-DOT. However, parameter recovery in DOT is a computationally expensive algorithm, especially when FD-HD measurements are required over a large and complex volume, as in the case of brain functional imaging. Therefore, computational tools for the light propagation modelling, known as the forward model, and the parameter recovery, known as the inverse problem, have been developed, in order to enable FD-HD-DOT. The forward model, within a diffusion approximation-based finite-element modelling framework, is accelerated by employing parallelization. A 10-fold speed increase when GPU architectures are available is achieved while maintaining high accuracy. For a very high-resolution finite-element model of the adult human head with ∼600,000 nodes, light propagation can be calculated at ∼0.25s per excitation source. Additionally, a framework for the sparse formulation of the inverse model, incorporating parallel computing, is proposed, achieving a 10-fold speed increase and a 100-fold memory efficiency, whilst maintaining reconstruction quality. Finally, to evaluate image reconstruction with and without the additional phase information, point spread functions have been simulated across a whole-scalp field of view in 24 subject-specific anatomical models using an experimentally derived noise model. The addition of phase information has shown to improve the image quality by reducing localization error by up to 59%, effective resolution by up to 21%, and depth penetration up to 5mm, as compared to using the intensity attenuation measurements alone. In addition, experimental data collected during a retinotopic experiment reveal that the phase data contains unique information about brain activity and enables images to be resolved for deeper brain regions.","abstract_has_math":false,"creators":["Doulgerakis, Matthaios"],"institution":"University of Birmingham","degree_name":"d_ph","degree_level":"d_ph","degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2021,"date_issued":"2021-07","date_published":"2021-07","updated_at":"2026-11-19T09:31:28Z","subjects":["QA75 Electronic computers. Computer science"],"languages":["eng"],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier.grantnumber","label":"Dc Identifier Grantnumber","values":["R01NS090874-08","RO1-CA132750","K01MH103594","R21MH109775"],"render_values":[{"text":"R01NS090874-08","href":null,"code":true},{"text":"RO1-CA132750","href":null,"code":true},{"text":"K01MH103594","href":null,"code":true},{"text":"R21MH109775","href":null,"code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":null,"outbound_label":null,"outbound_source":null,"record_url":null,"record_label":null,"record_kind":null,"record_source":null,"download_url":"http://etheses.bham.ac.uk//id/eprint/11672/7/Doulgerakis2021PhD.pdf","download_label":"Download PDF","download_source":"dc:identifier.uri","access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://etheses.bham.ac.uk/cgi/oai2?verb=GetRecord&metadataPrefix=uketd_dc&identifier=oai%3Aetheses.bham.ac.uk%3A11672","prefix":"uketd_dc"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.sponsor","label":"Sponsor","values":["nihr"]},{"key":"dc:creator","label":"Author","values":["Doulgerakis, Matthaios"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2021-07-21"]},{"key":"dc:date.issued","label":"Date","values":["2021-07"]},{"key":"dc:publisher.department","label":"Dc Publisher Department","values":["College of Engineering & Physical Sciences","School of Computer Science"]},{"key":"dc:publisher.institution","label":"Dc Publisher Institution","values":["University of Birmingham"]},{"key":"dc:relation.isreferencedby","label":"Dc Relation Isreferencedby","values":["http://etheses.bham.ac.uk//id/eprint/11672/"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"dc:type.qualificationlevel","label":"Dc Type Qualificationlevel","values":["d_ph"]},{"key":"dc:type.qualificationname","label":"Dc Type Qualificationname","values":["d_ph"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["QA75 Electronic computers. Computer science"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.dcterms_iso639_2","label":"Language (ISO 639-2)","values":["eng"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.grantnumber","label":"Dc Identifier Grantnumber","values":["R01NS090874-08","RO1-CA132750","K01MH103594","R21MH109775"]},{"key":"dc:identifier.uri","label":"Identifier URI","values":["http://etheses.bham.ac.uk//id/eprint/11672/7/Doulgerakis2021PhD.pdf"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Measurements of dynamic near-infrared (NIR) light attenuation across the human head together with model-based image reconstruction algorithms allow the recovery of three-dimensional spatial brain activation maps. Previous studies using high-density diffuse optical tomography (HD-DOT) systems have reported improved image quality over sparse arrays. Modulated NIR light, known as Frequency Domain (FD) NIR, enables measurements of phase shift along with amplitude attenuation. It is hypothesised that the utilization of these two sets of complementary data (phase and amplitude) for brain activity detection will result in an improvement in reconstructed image quality within HD-DOT. However, parameter recovery in DOT is a computationally expensive algorithm, especially when FD-HD measurements are required over a large and complex volume, as in the case of brain functional imaging. Therefore, computational tools for the light propagation modelling, known as the forward model, and the parameter recovery, known as the inverse problem, have been developed, in order to enable FD-HD-DOT. The forward model, within a diffusion approximation-based finite-element modelling framework, is accelerated by employing parallelization. A 10-fold speed increase when GPU architectures are available is achieved while maintaining high accuracy. For a very high-resolution finite-element model of the adult human head with ∼600,000 nodes, light propagation can be calculated at ∼0.25s per excitation source. Additionally, a framework for the sparse formulation of the inverse model, incorporating parallel computing, is proposed, achieving a 10-fold speed increase and a 100-fold memory efficiency, whilst maintaining reconstruction quality. Finally, to evaluate image reconstruction with and without the additional phase information, point spread functions have been simulated across a whole-scalp field of view in 24 subject-specific anatomical models using an experimentally derived noise model. The addition of phase information has shown to improve the image quality by reducing localization error by up to 59%, effective resolution by up to 21%, and depth penetration up to 5mm, as compared to using the intensity attenuation measurements alone. In addition, experimental data collected during a retinotopic experiment reveal that the phase data contains unique information about brain activity and enables images to be resolved for deeper brain regions."]},{"key":"dc:format","label":"Dc Format","values":["text"]},{"key":"dcterms:accessrights","label":"Dcterms Accessrights","values":["public"]},{"key":"dc:title","label":"Title","values":["Frequency domain high density diffuse optical tomography for functional brain imaging"]}]}],"canonical_facts":{"dc:contributor.sponsor":["nihr"],"dc:creator":["Doulgerakis, Matthaios"],"dc:date":["2021-07-21"],"dc:date.issued":["2021-07"],"dc:description.abstract":["Measurements of dynamic near-infrared (NIR) light attenuation across the human head together with model-based image reconstruction algorithms allow the recovery of three-dimensional spatial brain activation maps. Previous studies using high-density diffuse optical tomography (HD-DOT) systems have reported improved image quality over sparse arrays. Modulated NIR light, known as Frequency Domain (FD) NIR, enables measurements of phase shift along with amplitude attenuation. It is hypothesised that the utilization of these two sets of complementary data (phase and amplitude) for brain activity detection will result in an improvement in reconstructed image quality within HD-DOT. However, parameter recovery in DOT is a computationally expensive algorithm, especially when FD-HD measurements are required over a large and complex volume, as in the case of brain functional imaging. Therefore, computational tools for the light propagation modelling, known as the forward model, and the parameter recovery, known as the inverse problem, have been developed, in order to enable FD-HD-DOT. The forward model, within a diffusion approximation-based finite-element modelling framework, is accelerated by employing parallelization. A 10-fold speed increase when GPU architectures are available is achieved while maintaining high accuracy. For a very high-resolution finite-element model of the adult human head with ∼600,000 nodes, light propagation can be calculated at ∼0.25s per excitation source. Additionally, a framework for the sparse formulation of the inverse model, incorporating parallel computing, is proposed, achieving a 10-fold speed increase and a 100-fold memory efficiency, whilst maintaining reconstruction quality. Finally, to evaluate image reconstruction with and without the additional phase information, point spread functions have been simulated across a whole-scalp field of view in 24 subject-specific anatomical models using an experimentally derived noise model. The addition of phase information has shown to improve the image quality by reducing localization error by up to 59%, effective resolution by up to 21%, and depth penetration up to 5mm, as compared to using the intensity attenuation measurements alone. In addition, experimental data collected during a retinotopic experiment reveal that the phase data contains unique information about brain activity and enables images to be resolved for deeper brain regions."],"dc:format":["text"],"dc:identifier.grantnumber":["R01NS090874-08","RO1-CA132750","K01MH103594","R21MH109775"],"dc:identifier.uri":["http://etheses.bham.ac.uk//id/eprint/11672/7/Doulgerakis2021PhD.pdf"],"dc:language.dcterms_iso639_2":["eng"],"dc:publisher.department":["College of Engineering & Physical Sciences","School of Computer Science"],"dc:publisher.institution":["University of Birmingham"],"dc:relation.isreferencedby":["http://etheses.bham.ac.uk//id/eprint/11672/"],"dc:subject":["QA75 Electronic computers. Computer science"],"dcterms:accessrights":["public"],"dc:title":["Frequency domain high density diffuse optical tomography for functional brain imaging"],"dc:type":["Thesis"],"dc:type.qualificationlevel":["d_ph"],"dc:type.qualificationname":["d_ph"]},"updated_at":"2026-11-19T09:31:28Z"}