Tools and Resources for Large-Scale Morphometrics

Devine, Jay

Tools and Resources for Large-Scale Morphometrics

dc.contributor.advisor	Hallgrimsson, Benedikt
dc.contributor.author	Devine, Jay
dc.contributor.committeemember	Gonzalez, Paula
dc.contributor.committeemember	Percival, Christopher
dc.contributor.committeemember	Kurki, Helen
dc.contributor.committeemember	Epp, Jonathan
dc.date	2023-11
dc.date.accessioned	2023-08-09T17:00:09Z
dc.date.available	2023-08-09T17:00:09Z
dc.date.issued	2023-08
dc.description.abstract	Recent advances in imaging and machine learning have revolutionized both the acquisition and analysis of high-dimensional data. Unfortunately, these techniques have not been effectively exported to morphometrics, or the quantitative analysis of shape and form, which is central to biology and related disciplines for understanding phenotypic variation. Lack of implementations, different model systems and experimental designs, small sample sizes, and limited expertise continue to yield self-contained morphometric studies with unstandardized datasets that cannot be aggregated for increasingly powerful analyses. This dissertation presents heuristic and learning-based tools alongside big data resources for large-scale morphometrics. I begin by summarizing key morphometric paradigms, focusing on data acquisition, aggregation, and classification. Next, I develop an automated landmarking and shape optimization pipeline based on image registration and artificial neural networks. After demonstrating its efficacy, I use this phenotyping framework to build MusMorph, a database of mouse morphology data (N=10,056) containing anatomical atlases, aligned micro-computed tomography images, landmark configurations, and segmentations spanning numerous strain/genotype combinations and developmental stages. Finally, I introduce the R package pheble to perform a meta-analysis of learning-based classification algorithms across high-dimensional phenotypic datasets ranging in organismal family and anatomy. My analysis shows how ensemble learning, or combining predictions from individual learners, can maximize classification performance whilst being data agnostic. Altogether, these studies represent tools and resources that can accelerate, standardize, and augment morphometric analyses for novel phenomic inquiries.
dc.identifier.citation	Devine, J. (2023). Tools and resources for large-scale morphometrics (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.
dc.identifier.uri	https://hdl.handle.net/1880/116842
dc.identifier.uri	https://dx.doi.org/10.11575/PRISM/41684
dc.language.iso	en
dc.publisher.faculty	Arts
dc.publisher.institution	University of Calgary
dc.rights	University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.
dc.subject	classification
dc.subject	image registration
dc.subject	landmark
dc.subject	machine learning
dc.subject	morphometrics
dc.subject	phenomics
dc.subject.classification	Anatomy
dc.subject.classification	Genetics
dc.subject.classification	Artificial Intelligence
dc.subject.classification	Statistics
dc.subject.classification	Biophysics--Medical
dc.subject.classification	Radiology
dc.title	Tools and Resources for Large-Scale Morphometrics
dc.type	doctoral thesis
thesis.degree.discipline	Medicine – Medical Sciences
thesis.degree.grantor	University of Calgary
thesis.degree.name	Doctor of Philosophy (PhD)
ucalgary.thesis.accesssetbystudent	I do not require a thesis withhold – my thesis will have open access and can be viewed and downloaded publicly as soon as possible.