{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/122093"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/122093","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Maximizing utility of clinical trial neuroimaging data using advances in reproducible processing pipelines - benefits of modern diffusion reconstruction and structural connectivity methods","abstract":"Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2025-12-01","abstract_html":"Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2025-12-01","abstract_has_math":false,"creators":["Camacho, Paul B"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"Ph.D.","degree_level":"Dissertation","degree_discipline":"Neuroscience","degree_department":null,"school":null,"contributors":["Sutton, Brad P","Barbey, Aron K","Sadaghiani, Sepideh","Hernandez, Manuel E"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2023,"date_issued":"2023-12","date_published":"2023-12","updated_at":"2026-07-22T22:25:00Z","subjects":["Diffusion Weighted Imaging","Structural Connectivity","Resting-state Functional Connectivity","Healthy Aging","Cardio-respiratory Fitness","Motor Function"],"languages":["en","eng"],"rights":["Copyright 2023 Paul Camacho"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/2142/122093","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Sutton, Brad P","Barbey, Aron K","Sadaghiani, Sepideh","Hernandez, Manuel E"]},{"key":"dc:creator","label":"Author","values":["Camacho, Paul B"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2023-12","2023-09-07"]},{"key":"dc:type","label":"Dc Type","values":["text"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Neuroscience"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Ph.D."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois at Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Diffusion Weighted Imaging","Structural Connectivity","Resting-state Functional Connectivity","Healthy Aging","Cardio-respiratory Fitness","Motor Function"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en","eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2023 Paul Camacho"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://hdl.handle.net/2142/122093"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2025-12-01","The student, Paul Camacho, accepted the attached license on 2023-08-24 at 08:13.","The student, Paul Camacho, submitted this Dissertation for approval on 2023-08-25 at 14:42.","This Dissertation was approved for publication on 2023-09-07 at 10:43.","DSpace SAF Submission Ingestion Package generated from Vireo submission #19814 on 2024-03-01 at 13:29:20","The variety of options available for preprocessing and analyses of MRI data are a major area of concern in reproducible neuroscience. The level of potential variance in findings and clinical interpretations of data associated with these choices is especially important for the translation of findings from studies using cutting-edge MRI methods to use in a clinical context. Clinical trials - especially in the pilot phase - are typically under-powered and not designed to leverage cutting edge analysis tools as they investigate well-posed research questions. Furthermore, adoption of processing software in a clinical trials or large-scale (i.e. consortia-level studies) setting requires feasible times for processing to yield actionable results in a reasonable amount of time. The goal of this research is to determine the effects of clinical trial-feasible MRI data processing and analyses methods on the predictive power of connectivity-based biomarkers sought in clinical and translational neuroscience studies. With the advent of large-scale data collection projects, such as the human connectome project, UK Biobank, and other national-level efforts, there has developed a strong push for the development of reproducible processing pipelines to extract compatible results from large data sets. These tools have enabled automated or semi-automated processing of data with complex analysis pathways that are scalable, reproducible, and well documented. This does more than make average neuroimaging users into sophisticated high performance computer users. It enables them to run significantly more complex workflows and seamlessly link advanced anatomical, diffusion, and functional imaging pipelines together. A significant gap exists though, in facilitating the set up of coupled processing pipelines and in demonstrating the power and benefits of applying these to our common neuroimaging studies. In this work, I will further develop scripts and best practices in setting up a high performance computing environment leveraging reproducible neuroimaging software. Further, I will demonstrate the significant gains in resting-state functional connectivity prediction accuracy and functional fitness modeling that can be had by applying advanced diffusion models and structural connectivity generation workflows to clinical trial diffusion magnetic resonance imaging data. Validation of the effects of exercise or physical therapy interventions on the brain relies on spatially localized quantitative metrics provided by neuroimaging data. In the absence of a ground truth wiring diagram of the brain to explain function and response to insult, these image metrics describe structural properties or activity in their respective spatial and temporal sampling windows. Although each sampling unit in neuroimaging inherently summarizes an array of cells, tissues, and structures, changes to metrics describing that unit - and realistically other sampling units of the complex system that is the brain - allow for biomarkers of change to the brain. Between persons, significant biomarkers that predict clinical dysfunction inform our mechanistic understanding of the brain. Models that capture the most relevant features of their respective image should yield more robust and responsive biomarkers that hold stronger predictive value of clinical dysfunction. While many models are seen in the literature and produce metrics used as in vivo proxies of brain tissue properties or connectivity, there is an apparent need for models that are better informed by the underlying physiology of the brain. In an effort to produce better biomarkers of motor function and cardio-respiratory fitness, here we compare standard diffusion tensor imaging to more recent diffusion reconstruction methods informed by different assumptions that have been shown to better characterize aspects of structural connectivity poorly captured by the standard approach. The main hypothesis of this work is that structural connectivity pipelines that better reflect the underlying tissue will better predict functional connectivity and function, capturing more information through the simulation of indirect connections."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Maximizing utility of clinical trial neuroimaging data using advances in reproducible processing pipelines - benefits of modern diffusion reconstruction and structural connectivity methods"]}]}],"canonical_facts":{"dc:contributor":["Sutton, Brad P","Barbey, Aron K","Sadaghiani, Sepideh","Hernandez, Manuel E"],"dc:creator":["Camacho, Paul B"],"dc:date":["2023-12","2023-09-07"],"dc:description":["Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2025-12-01","The student, Paul Camacho, accepted the attached license on 2023-08-24 at 08:13.","The student, Paul Camacho, submitted this Dissertation for approval on 2023-08-25 at 14:42.","This Dissertation was approved for publication on 2023-09-07 at 10:43.","DSpace SAF Submission Ingestion Package generated from Vireo submission #19814 on 2024-03-01 at 13:29:20","The variety of options available for preprocessing and analyses of MRI data are a major area of concern in reproducible neuroscience. The level of potential variance in findings and clinical interpretations of data associated with these choices is especially important for the translation of findings from studies using cutting-edge MRI methods to use in a clinical context. Clinical trials - especially in the pilot phase - are typically under-powered and not designed to leverage cutting edge analysis tools as they investigate well-posed research questions. Furthermore, adoption of processing software in a clinical trials or large-scale (i.e. consortia-level studies) setting requires feasible times for processing to yield actionable results in a reasonable amount of time. The goal of this research is to determine the effects of clinical trial-feasible MRI data processing and analyses methods on the predictive power of connectivity-based biomarkers sought in clinical and translational neuroscience studies. With the advent of large-scale data collection projects, such as the human connectome project, UK Biobank, and other national-level efforts, there has developed a strong push for the development of reproducible processing pipelines to extract compatible results from large data sets. These tools have enabled automated or semi-automated processing of data with complex analysis pathways that are scalable, reproducible, and well documented. This does more than make average neuroimaging users into sophisticated high performance computer users. It enables them to run significantly more complex workflows and seamlessly link advanced anatomical, diffusion, and functional imaging pipelines together. A significant gap exists though, in facilitating the set up of coupled processing pipelines and in demonstrating the power and benefits of applying these to our common neuroimaging studies. In this work, I will further develop scripts and best practices in setting up a high performance computing environment leveraging reproducible neuroimaging software. Further, I will demonstrate the significant gains in resting-state functional connectivity prediction accuracy and functional fitness modeling that can be had by applying advanced diffusion models and structural connectivity generation workflows to clinical trial diffusion magnetic resonance imaging data. Validation of the effects of exercise or physical therapy interventions on the brain relies on spatially localized quantitative metrics provided by neuroimaging data. In the absence of a ground truth wiring diagram of the brain to explain function and response to insult, these image metrics describe structural properties or activity in their respective spatial and temporal sampling windows. Although each sampling unit in neuroimaging inherently summarizes an array of cells, tissues, and structures, changes to metrics describing that unit - and realistically other sampling units of the complex system that is the brain - allow for biomarkers of change to the brain. Between persons, significant biomarkers that predict clinical dysfunction inform our mechanistic understanding of the brain. Models that capture the most relevant features of their respective image should yield more robust and responsive biomarkers that hold stronger predictive value of clinical dysfunction. While many models are seen in the literature and produce metrics used as in vivo proxies of brain tissue properties or connectivity, there is an apparent need for models that are better informed by the underlying physiology of the brain. In an effort to produce better biomarkers of motor function and cardio-respiratory fitness, here we compare standard diffusion tensor imaging to more recent diffusion reconstruction methods informed by different assumptions that have been shown to better characterize aspects of structural connectivity poorly captured by the standard approach. The main hypothesis of this work is that structural connectivity pipelines that better reflect the underlying tissue will better predict functional connectivity and function, capturing more information through the simulation of indirect connections."],"dc:format":["application/pdf"],"dc:identifier":["https://hdl.handle.net/2142/122093"],"dc:language":["en","eng"],"dc:rights":["Copyright 2023 Paul Camacho"],"dc:subject":["Diffusion Weighted Imaging","Structural Connectivity","Resting-state Functional Connectivity","Healthy Aging","Cardio-respiratory Fitness","Motor Function"],"dc:title":["Maximizing utility of clinical trial neuroimaging data using advances in reproducible processing pipelines - benefits of modern diffusion reconstruction and structural connectivity methods"],"dc:type":["text"],"thesis:degree_discipline":["Neuroscience"],"thesis:degree_level":["Dissertation"],"thesis:degree_name":["Ph.D."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:25:00Z"}