TubeTK/Vascular Pattern Analysis

From KitwarePublic
Jump to: navigation, search

Twin Study Involving Retinal Vessels

Background

  • "Toward automatic phenotyping of retinal images from genetically determined mono- and dizygotic twins using amplitude modulation-frequency modulation methods"
    • Peter Soliz, Herbert T. Davis III, VisionQuest Inc. (United States); Victor Murray, Marios Pattichis, The Univ. of New Mexico (United States); E. Simon Barriga, VisionQuest Biomedical, LLC (United States); Stephen R. Russell, The Univ. of Iowa Hospitals and Clinics
    • A new method for classifying retinal images based on features derived from the application of amplitude modulation-frequency modulation (AM-FM) methods is presented. Retinal images from identical and fraternal twins who presented with age-related macular degeneration (AMD) were processed to determine whether AM-FM could be used to differentiate between the two types of twins. Results of the automatic classifier agreed with the findings of other researchers in explaining the variation of the disease between the related twins. AM-FM features classified 72% of the twins correctly. Visual analysis found that genetics could explain between 46% and 71% of the variance.
    • SPIE 2010
  • "Heritable Features of the Optic Disc: A Novel Twin Method for Determining Genetic Significance"
  • "Automated Quantification of Inherited Phenotypes from Color Images: a Twin Study of the Variability of the Optic Nerve Head Shape."
    • Tang L, Scheetz TE, Mackey DA, Hewitt AW, Fingert J, Kwon YH, Quellec G, Reinhardt JM, Abramoff MD.
    • Ophthalmology and Visual Sciences, University of Iowa, Iowa City, United States.
    • http://www.ncbi.nlm.nih.gov/pubmed/20505201
  • "Retinal atlas statistics from color fundus images"
    • Sangyeol Lee, Michael D. Abramoff, Joseph M. Reinhardt, The Univ. of Iowa
    • An atlas provides a reference anatomic structure and an associated coordinate system. We present a method for generating an atlas of the human retina from 500 color fundus image pairs. Key retinal anatomic features are selected for atlas landmarks; disk center, fovea, and vessel main arcade. An atlas coordinate system is defined based on the statistics of the landmarks. Images from the population are warped into the atlas space to produce a statistical retinal atlas which can be used automatic diagnosis, concise indexing, semantic blending, etc.
    • SPIE 2010
  • "Automatic determination of the artery vein ratio in retinal images."
    • Meindert Niemeijer, Bram van Ginneken, Univ. Medical Ctr. Utrecht (Netherlands);
    • Michael D. Abramoff, The Univ. of Iowa Hospitals and Clinics (United States)
    • The ratio between the width of the arteries and veins (Arteriolar-to-Venular diameter Ratio, AVR) on the retina, is well established to be predictive of several cardiovascular conditions. This work presents an automatic method that detects the location of the optic disc, determines the appropriate region of interest (ROI), classifies the vessels in the ROI into arteries and veins, measures their widths and calculates the AVR. We test the system on a set of 15 images obtaining promising results when compared with a human observer who determined the AVR.
    • SPIE 2010
  • "Use of a twin dataset to identify AMD-related visual patterns controlled by genetic factors"
    • Gwenole Quellec, Michael D. Abramoff, Stephen R. Russell, The Univ. of Iowa Hospitals and Clinics
    • Genotype-to-phenotype mapping is expected to improve the diagnosis and treatment of AMD. In this study, we identified visual patterns related to AMD controlled by genetic factors, using a dataset of eye fundus photographs from 41 monozygotic or dizygotic twin pairs. We first developed a detector for hard and soft drusen. Then we defined forty visual features characterizing the detected drusen. A distance measure between twin pairs was defined for each visual feature, in order to differentiate monozygotic from dizygotic twins. The predictions of several visual features (65.9% accuracy) are comparable or better than the predictions of human experts.
    • SPIE 2010
  • "Retinal blood vessel tortuosity: a prospective diagnostic tool for the analysis of retinopathies"
    • Vinayak S. Joshi, Joseph M. Reinhardt, Michael D. Abramoff, The Univ. of Iowa
    • Retinal blood vessel network may provide information about occurrence and severity of retinopathies. Presence of retinopathies often results in vessel tortuosity. To estimate the tortuosity, we examine the shape of blood vessel by measuring angle of curvature, arc length of curved vessel over its chord length, number of turn sign changes and define a tortuosity index (TI) to characterize normal or abnormal vessel. We applied the method to 15 patients with FSHD, producing the same rank-ordered list of vessel tortuosity as produced with the assessment by 4 physicians. Accordingly, TI may be a useful tool in the diagnosis of retinopathies.

Intra-cranial Vessel Patterns of Populations

Lung Vessel Patterns

  • "Human airway tree structure query atlas"
    • Gary E. Christensen, Nathan Burnette, Weichen Gao, Matineh Shaker, Joseph M.

Reinhardt, Janice E. Cook-Granroth, Geoffrey McLennan M.D., Eric A. Hoffman

    • The Univ. of Iowa
    • SPIE 2010
    • A queryable electronic atlas was developed to quantitatively characterize the normal human lung

airway tree and to provide a better understanding of the lung for diagnosing diseases and evaluating treatments. The atlas consists of more than 3000 measurements of the airway tree for each scan. The atlas provides user friendly interfaces for interrogating population statistics, comparing populations, comparing individuals to populations, and comparing individuals to other individuals. Populations can be selected based on age, gender, race, ethnicity, and normalcy/disease. The atlas database consists of measurements from 129 individuals mainly in the age ranges 20-30 and 40-50 years old.