Share Email Print
cover

Proceedings Paper

Cascaded classifier for large-scale data applied to automatic segmentation of articular cartilage
Author(s): Adhish Prasoon; Christian Igel; Marco Loog; François Lauze; Erik Dam; Mads Nielsen
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Many classification/segmentation tasks in medical imaging are particularly challenging for machine learning algorithms because of the huge amount of training data required to cover biological variability. Learning methods scaling badly in the number of training data points may not be applicable. This may exclude powerful classifiers with good generalization performance such as standard non-linear support vector machines (SVMs). Further, many medical imaging problems have highly imbalanced class populations, because the object to be segmented has only few pixels/voxels compared to the background. This article presents a two-stage classifier for large-scale medical imaging problems. In the first stage, a classifier that is easily trainable on large data sets is employed. The class imbalance is exploited and the classifier is adjusted to correctly detect background with a very high accuracy. Only the comparatively few data points not identified as background are passed to the second stage. Here a powerful classifier with high training time complexity can be employed for making the final decision whether a data point belongs to the object or not. We applied our method to the problem of automatically segmenting tibial articular cartilage from knee MRI scans. We show that by using nearest neighbor (kNN) in the first stage we can reduce the amount of data for training a non-linear SVM in the second stage. The cascaded system achieves better results than the state-of-the-art method relying on a single kNN classifier.

Paper Details

Date Published: 24 February 2012
PDF: 9 pages
Proc. SPIE 8314, Medical Imaging 2012: Image Processing, 83144V (24 February 2012); doi: 10.1117/12.910809
Show Author Affiliations
Adhish Prasoon, Univ. of Copenhagen (Denmark)
Christian Igel, Univ. of Copenhagen (Denmark)
Marco Loog, Univ. of Copenhagen (Denmark)
Delft Univ. of Technology (Netherlands)
François Lauze, Univ. of Copenhagen (Denmark)
Erik Dam, BioMed IQ (Denmark)
Mads Nielsen, Univ. of Copenhagen (Denmark)
BioMed IQ (Denmark)


Published in SPIE Proceedings Vol. 8314:
Medical Imaging 2012: Image Processing
David R. Haynor; Sébastien Ourselin, Editor(s)

© SPIE. Terms of Use
Back to Top