Pathology detection on medical images based in oriented active appearance models

Xinjian Chen; Jayaram K. Udupa; Abass Alavi; Drew A. Torigian

doi:10.1117/12.844543

9 March 2010 Pathology detection on medical images based in oriented active appearance models

Xinjian Chen, Jayaram K. Udupa, Abass Alavi, Drew A. Torigian

Proceedings Volume 7624, Medical Imaging 2010: Computer-Aided Diagnosis; 76243M (2010) https://doi.org/10.1117/12.844543
Event: SPIE Medical Imaging, 2010, San Diego, California, United States

Abstract

In this paper, we propose a novel, general paradigm based on creating a statistical geographic model of shape and appearance of normal body regions. Any deviations from the normality information captured in a given patient image are highlighted and expressed as a fuzzy pathology image. We study the feasibility of this idea in 2D images via Oriented Active Appearance Models (OAAM). The OAAM synergistically combines AAM and live-wire concepts. The approach consists of three main stages: model building, segmentation, and pathology detection. The model is built on image data from normal subjects. The model currently includes shape and texture information. A variety of other information (functional, morphometric) can be added in the future. For segmentation, a novel automatic object recognition method is proposed which strategically combines the AAM with the live-wire method. A two level dynamic programming method is used to do the finer delineation. During the process of segmentation, a multi-object strategy is used for improving recognition and delineation accuracy. For pathology detection, the model is first fit to the given image as best as possible via recognition and delineation of the objects included in the model. Subsequently, a fuzzy pathology image is generated that expresses deviations in appearance of the given image form the texture information contained in the model. The proposed method was tested on two clinical CT medical image datasets each consisting of 40 images. Our preliminary results indicate high segmentation accuracy (TPVF>97%, FPVF<0.5%) for delineating objects by the multi-object strategy with good pathology detection results suggesting the feasibility of the proposed system.

Citation Download Citation

Xinjian Chen, Jayaram K. Udupa, Abass Alavi, and Drew A. Torigian "Pathology detection on medical images based in oriented active appearance models", Proc. SPIE 7624, Medical Imaging 2010: Computer-Aided Diagnosis, 76243M (9 March 2010); https://doi.org/10.1117/12.844543

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
8 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Image segmentation

Pathology

Computed tomography

Chest

Medical imaging

Fuzzy logic

Data modeling

Show All Keywords

Keywords/Phrases

Search In:

Publication Years