Statistical shape analysis of automatically segmented femur bones: Data from the osteoarthritis initiative

Satyananda Kashyap; Ipek Oguz; Milan Sonka

doi:10.1109/ISBI.2016.7493351

Back

Conference proceeding

Statistical shape analysis of automatically segmented femur bones: Data from the osteoarthritis initiative

Satyananda Kashyap, Ipek Oguz and Milan Sonka

2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI), Vol.2016-, pp.651-655

04/2016

DOI: 10.1109/ISBI.2016.7493351

View Online

Abstract

Early detection of structural differences associated with the osteoarthritis (OA) of the knee is crucial for enabling effective clinical trials for testing potential early interventional disease modifying drugs. Highly localized quantification methods are needed for assessing differences between patient populations. In this paper we present a fully automated method to quantify femoral bone shape differences between subjects having progressive and non-progressive osteoarthritis (OA) from MRI scans of the knee. The bone is identified using a fully automated approach based on the layered optimal graph image segmentation of multiple objects and surfaces (LOGISMOS). Statistical shape analysis is performed using the spherical harmonics based point distribution model (SPHARM-PDM). Data from the Osteoarthritis Initiative (OAI) for subjects with a Kellgren Lawrence (KL) grade of 2 at baseline were used in the study. Shape differences between the progressor and non-progressor subject groups were compared at baseline, 1-year and 2-year followups. 576 MRI scans in total were analyzed. We found significant differences (p < 0.05) in the bone shape between the progressor versus non-progressor populations at the 1-year and 2-year follow up visits, with the most pronounced shape differences observed around the trochlear groove region.

Harmonic Analysis

Magnetic Resonance Imaging

Osteoarthritis

Sociology

Statistics

Bones

Cost function

Graph Search

Knee MRI

LOGISMOS

Shape

Shape Analysis

SPHARM-PDM

Trochlear groove

Details

Title: Subtitle: Statistical shape analysis of automatically segmented femur bones: Data from the osteoarthritis initiative
Creators: Satyananda Kashyap - University of Iowa
Ipek Oguz - University of Iowa
Milan Sonka - University of Iowa
Resource Type: Conference proceeding
Publication Details: 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI), Vol.2016-, pp.651-655
DOI: 10.1109/ISBI.2016.7493351
ISSN: 1945-7928
eISSN: 1945-8452
Publisher: IEEE
Language: English
Date published: 04/2016
Academic Unit: Roy J. Carver Department of Biomedical Engineering; Electrical and Computer Engineering; Psychiatry; Radiation Oncology; Fraternal Order of Eagles Diabetes Research Center; Injury Prevention Research Center; Ophthalmology and Visual Sciences
Record Identifier: 9984186692702771

Metrics

14 Record Views