High-dimensional pattern regression using machine learning

From medical images to continuous clinical variables

Ying Wang, Yong Fan, Priyanka Bhatt, Christos Davatzikos

Research output: Contribution to journalArticle

Abstract

This paper presents a general methodology for high-dimensional pattern regression on medical images via machine learning techniques. Compared with pattern classification studies, pattern regression considers the problem of estimating continuous rather than categorical variables, and can be more challenging. It is also clinically important, since it can be used to estimate disease stage and predict clinical progression from images. In this work, adaptive regional feature extraction approach is used along with other common feature extraction methods, and feature selection technique is adopted to produce a small number of discriminative features for optimal regression performance. Then the Relevance Vector Machine (RVM) is used to build regression models based on selected features. To get stable regression models from limited training samples, a bagging framework is adopted to build ensemble basis regressors derived from multiple bootstrap training samples, and thus to alleviate the effects of outliers as well as facilitate the optimal model parameter selection. Finally, this regression scheme is tested on simulated data and real data via cross-validation. Experimental results demonstrate that this regression scheme achieves higher estimation accuracy and better generalizing ability than Support Vector Regression (SVR).

Original languageEnglish (US)
Pages (from-to)1519-1535
Number of pages17
JournalNeuroImage
Volume50
Issue number4
DOIs
StatePublished - May 1 2010
Externally publishedYes

Fingerprint

Machine Learning

Keywords

  • Adaptive regional clustering
  • Alzheimer's disease
  • High-dimensionality pattern regression
  • MRI
  • Relevance vector regression

ASJC Scopus subject areas

  • Cognitive Neuroscience
  • Neurology

Cite this

High-dimensional pattern regression using machine learning : From medical images to continuous clinical variables. / Wang, Ying; Fan, Yong; Bhatt, Priyanka; Davatzikos, Christos.

In: NeuroImage, Vol. 50, No. 4, 01.05.2010, p. 1519-1535.

Research output: Contribution to journalArticle

Wang, Ying ; Fan, Yong ; Bhatt, Priyanka ; Davatzikos, Christos. / High-dimensional pattern regression using machine learning : From medical images to continuous clinical variables. In: NeuroImage. 2010 ; Vol. 50, No. 4. pp. 1519-1535.
@article{2840461668a74042beeccae8ef9b46c0,
title = "High-dimensional pattern regression using machine learning: From medical images to continuous clinical variables",
abstract = "This paper presents a general methodology for high-dimensional pattern regression on medical images via machine learning techniques. Compared with pattern classification studies, pattern regression considers the problem of estimating continuous rather than categorical variables, and can be more challenging. It is also clinically important, since it can be used to estimate disease stage and predict clinical progression from images. In this work, adaptive regional feature extraction approach is used along with other common feature extraction methods, and feature selection technique is adopted to produce a small number of discriminative features for optimal regression performance. Then the Relevance Vector Machine (RVM) is used to build regression models based on selected features. To get stable regression models from limited training samples, a bagging framework is adopted to build ensemble basis regressors derived from multiple bootstrap training samples, and thus to alleviate the effects of outliers as well as facilitate the optimal model parameter selection. Finally, this regression scheme is tested on simulated data and real data via cross-validation. Experimental results demonstrate that this regression scheme achieves higher estimation accuracy and better generalizing ability than Support Vector Regression (SVR).",
keywords = "Adaptive regional clustering, Alzheimer's disease, High-dimensionality pattern regression, MRI, Relevance vector regression",
author = "Ying Wang and Yong Fan and Priyanka Bhatt and Christos Davatzikos",
year = "2010",
month = "5",
day = "1",
doi = "10.1016/j.neuroimage.2009.12.092",
language = "English (US)",
volume = "50",
pages = "1519--1535",
journal = "NeuroImage",
issn = "1053-8119",
publisher = "Academic Press Inc.",
number = "4",

}

TY - JOUR

T1 - High-dimensional pattern regression using machine learning

T2 - From medical images to continuous clinical variables

AU - Wang, Ying

AU - Fan, Yong

AU - Bhatt, Priyanka

AU - Davatzikos, Christos

PY - 2010/5/1

Y1 - 2010/5/1

N2 - This paper presents a general methodology for high-dimensional pattern regression on medical images via machine learning techniques. Compared with pattern classification studies, pattern regression considers the problem of estimating continuous rather than categorical variables, and can be more challenging. It is also clinically important, since it can be used to estimate disease stage and predict clinical progression from images. In this work, adaptive regional feature extraction approach is used along with other common feature extraction methods, and feature selection technique is adopted to produce a small number of discriminative features for optimal regression performance. Then the Relevance Vector Machine (RVM) is used to build regression models based on selected features. To get stable regression models from limited training samples, a bagging framework is adopted to build ensemble basis regressors derived from multiple bootstrap training samples, and thus to alleviate the effects of outliers as well as facilitate the optimal model parameter selection. Finally, this regression scheme is tested on simulated data and real data via cross-validation. Experimental results demonstrate that this regression scheme achieves higher estimation accuracy and better generalizing ability than Support Vector Regression (SVR).

AB - This paper presents a general methodology for high-dimensional pattern regression on medical images via machine learning techniques. Compared with pattern classification studies, pattern regression considers the problem of estimating continuous rather than categorical variables, and can be more challenging. It is also clinically important, since it can be used to estimate disease stage and predict clinical progression from images. In this work, adaptive regional feature extraction approach is used along with other common feature extraction methods, and feature selection technique is adopted to produce a small number of discriminative features for optimal regression performance. Then the Relevance Vector Machine (RVM) is used to build regression models based on selected features. To get stable regression models from limited training samples, a bagging framework is adopted to build ensemble basis regressors derived from multiple bootstrap training samples, and thus to alleviate the effects of outliers as well as facilitate the optimal model parameter selection. Finally, this regression scheme is tested on simulated data and real data via cross-validation. Experimental results demonstrate that this regression scheme achieves higher estimation accuracy and better generalizing ability than Support Vector Regression (SVR).

KW - Adaptive regional clustering

KW - Alzheimer's disease

KW - High-dimensionality pattern regression

KW - MRI

KW - Relevance vector regression

UR - http://www.scopus.com/inward/record.url?scp=77950588524&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77950588524&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2009.12.092

DO - 10.1016/j.neuroimage.2009.12.092

M3 - Article

VL - 50

SP - 1519

EP - 1535

JO - NeuroImage

JF - NeuroImage

SN - 1053-8119

IS - 4

ER -