Accurate estimation of retinal vessel width using bagged decision trees and an extended multiresolution Hermite model

Carmen Alina Lupaşcu, Domenico Tegolo, Emanuele Trucco

    Research output: Contribution to journalArticle

    29 Citations (Scopus)

    Abstract

    We present an algorithm estimating the width of retinal vessels in fundus camera images. The algorithm uses a novel parametric surface model of the cross-sectional intensities of vessels, and ensembles of bagged decision trees to estimate the local width from the parameters of the best-fit surface.We report comparative tests with REVIEW, currently the public database of reference for retinal width estimation, containing 16 images with 193 annotated vessel segments and 5066 profile points annotated manually by three independent experts. Comparative tests are reported also with our own set of 378 vessel widths selected sparsely in 38 images from the Tayside Scotland diabetic retinopathy screening programme and annotated manually by two clinicians. We obtain considerably better accuracies compared to leading methods in REVIEW tests and in Tayside tests.An important advantage of our method is its stability (success rate, i.e., meaningful measurement returned, of 100% on all REVIEW data sets and on the Tayside data set) compared to a variety of methods from the literature. We also find that results depend crucially on testing data and conditions, and discuss criteria for selecting a training set yielding optimal accuracy.
    Original languageEnglish
    Pages (from-to)1164-1180
    Number of pages17
    JournalMedical Image Analysis
    Volume17
    Issue number8
    DOIs
    Publication statusPublished - 1 Dec 2013

    Fingerprint

    Retinal Vessels
    Decision Trees
    Decision trees
    Screening
    Cameras
    Scotland
    Diabetic Retinopathy
    Testing
    Databases
    Datasets

    Cite this

    @article{6d3b3ed71e9c48cc8d6ac2cca609439e,
    title = "Accurate estimation of retinal vessel width using bagged decision trees and an extended multiresolution Hermite model",
    abstract = "We present an algorithm estimating the width of retinal vessels in fundus camera images. The algorithm uses a novel parametric surface model of the cross-sectional intensities of vessels, and ensembles of bagged decision trees to estimate the local width from the parameters of the best-fit surface.We report comparative tests with REVIEW, currently the public database of reference for retinal width estimation, containing 16 images with 193 annotated vessel segments and 5066 profile points annotated manually by three independent experts. Comparative tests are reported also with our own set of 378 vessel widths selected sparsely in 38 images from the Tayside Scotland diabetic retinopathy screening programme and annotated manually by two clinicians. We obtain considerably better accuracies compared to leading methods in REVIEW tests and in Tayside tests.An important advantage of our method is its stability (success rate, i.e., meaningful measurement returned, of 100{\%} on all REVIEW data sets and on the Tayside data set) compared to a variety of methods from the literature. We also find that results depend crucially on testing data and conditions, and discuss criteria for selecting a training set yielding optimal accuracy.",
    author = "Lupaşcu, {Carmen Alina} and Domenico Tegolo and Emanuele Trucco",
    year = "2013",
    month = "12",
    day = "1",
    doi = "10.1016/j.media.2013.07.006",
    language = "English",
    volume = "17",
    pages = "1164--1180",
    journal = "Medical Image Analysis",
    issn = "1361-8415",
    publisher = "Elsevier",
    number = "8",

    }

    Accurate estimation of retinal vessel width using bagged decision trees and an extended multiresolution Hermite model. / Lupaşcu, Carmen Alina; Tegolo, Domenico; Trucco, Emanuele.

    In: Medical Image Analysis, Vol. 17, No. 8, 01.12.2013, p. 1164-1180.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - Accurate estimation of retinal vessel width using bagged decision trees and an extended multiresolution Hermite model

    AU - Lupaşcu, Carmen Alina

    AU - Tegolo, Domenico

    AU - Trucco, Emanuele

    PY - 2013/12/1

    Y1 - 2013/12/1

    N2 - We present an algorithm estimating the width of retinal vessels in fundus camera images. The algorithm uses a novel parametric surface model of the cross-sectional intensities of vessels, and ensembles of bagged decision trees to estimate the local width from the parameters of the best-fit surface.We report comparative tests with REVIEW, currently the public database of reference for retinal width estimation, containing 16 images with 193 annotated vessel segments and 5066 profile points annotated manually by three independent experts. Comparative tests are reported also with our own set of 378 vessel widths selected sparsely in 38 images from the Tayside Scotland diabetic retinopathy screening programme and annotated manually by two clinicians. We obtain considerably better accuracies compared to leading methods in REVIEW tests and in Tayside tests.An important advantage of our method is its stability (success rate, i.e., meaningful measurement returned, of 100% on all REVIEW data sets and on the Tayside data set) compared to a variety of methods from the literature. We also find that results depend crucially on testing data and conditions, and discuss criteria for selecting a training set yielding optimal accuracy.

    AB - We present an algorithm estimating the width of retinal vessels in fundus camera images. The algorithm uses a novel parametric surface model of the cross-sectional intensities of vessels, and ensembles of bagged decision trees to estimate the local width from the parameters of the best-fit surface.We report comparative tests with REVIEW, currently the public database of reference for retinal width estimation, containing 16 images with 193 annotated vessel segments and 5066 profile points annotated manually by three independent experts. Comparative tests are reported also with our own set of 378 vessel widths selected sparsely in 38 images from the Tayside Scotland diabetic retinopathy screening programme and annotated manually by two clinicians. We obtain considerably better accuracies compared to leading methods in REVIEW tests and in Tayside tests.An important advantage of our method is its stability (success rate, i.e., meaningful measurement returned, of 100% on all REVIEW data sets and on the Tayside data set) compared to a variety of methods from the literature. We also find that results depend crucially on testing data and conditions, and discuss criteria for selecting a training set yielding optimal accuracy.

    UR - http://www.scopus.com/inward/record.url?scp=84883283641&partnerID=8YFLogxK

    U2 - 10.1016/j.media.2013.07.006

    DO - 10.1016/j.media.2013.07.006

    M3 - Article

    VL - 17

    SP - 1164

    EP - 1180

    JO - Medical Image Analysis

    JF - Medical Image Analysis

    SN - 1361-8415

    IS - 8

    ER -