Eresen, Aydin (2018-08). Detection of MRI Biomarkers of Golden Retriever Duchenne Muscular Dystrophy. Doctoral Dissertation. Thesis uri icon

abstract

  • Golden retriever muscular dystrophy (GRMD) is a spontaneous X-linked canine model of Duchenne muscular dystrophy (DMD) with the affected animals developing a progressively fatal disease, similar to the human condition. As a genetically homologous animal model, GRMD has increasingly been used in natural history studies and studies assessing treatment outcome. There is a great demand for accurate outcome measures across all disease stages to improve the understanding of natural history and to facilitate clinical trials. Histology images are widely used for accurate outcome measures across all disease stages. With a highly invasive method as ground-truth, a variety of non-invasive methods are frequently assessed to extract information corresponding to biological characteristics. Due to high soft-tissue contrast images, MRI is commonly preferred imaging modality to assess GRMD. Spatial correspondence between histology and MRI is a critical step in the quantitative evaluation of skeletal muscle in GRMD. Registration becomes technically challenging due to non-orthogonal histology section orientation, section distortion, and the different image contrast and resolution. This research dissertation proposed a framework for accurate histology to MRI registration and textural analysis methods to describe non-invasive MRI biomarkers utilizing multi-sequence MRI images. The experiments showed that textural features of qualitative and quantitative MRI images can be reliably used for disease assessments and treatment monitoring.
  • Golden retriever muscular dystrophy (GRMD) is a spontaneous X-linked canine model of
    Duchenne muscular dystrophy (DMD) with the affected animals developing a progressively fatal
    disease, similar to the human condition. As a genetically homologous animal model, GRMD has
    increasingly been used in natural history studies and studies assessing treatment outcome. There
    is a great demand for accurate outcome measures across all disease stages to improve the
    understanding of natural history and to facilitate clinical trials. Histology images are widely used
    for accurate outcome measures across all disease stages. With a highly invasive method as ground-truth,
    a variety of non-invasive methods are frequently assessed to extract information
    corresponding to biological characteristics. Due to high soft-tissue contrast images, MRI is
    commonly preferred imaging modality to assess GRMD. Spatial correspondence between
    histology and MRI is a critical step in the quantitative evaluation of skeletal muscle in GRMD.
    Registration becomes technically challenging due to non-orthogonal histology section orientation,
    section distortion, and the different image contrast and resolution. This research dissertation
    proposed a framework for accurate histology to MRI registration and textural analysis methods to
    describe non-invasive MRI biomarkers utilizing multi-sequence MRI images. The experiments
    showed that textural features of qualitative and quantitative MRI images can be reliably used for
    disease assessments and treatment monitoring.

publication date

  • August 2018