Machine learning as a means toward precision diagnostics and prognostics

A. Sotiras, B. Gaonkar, H. Eavani, N. Honnorat, E. Varol, A. Dong, C. Davatzikos

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Machine learning plays an essential role in medical imaging. Pattern analysis techniques can identify, and quantify, subtle and spatially complex patterns of disease-induced changes in the brain despite confounding statistical noise and inter-individual variability. This allows the construction of sensitive biomarkers that can identify disease, or risk of developing it, and characterize future clinical progression on an individual patient basis. Thus pattern analysis techniques have become an indispensable part of the growing need for personalized, predictive medicine. However, despite important advances, several challenges remain before they can gain widespread acceptance as tools for precision diagnostics and prognostics in clinical practice. These include: (i) feature extraction and dimensionality reduction; (ii) readily interpreting complex multivariate models; and (iii) elucidating disease heterogeneity. In this chapter, we describe these challenges, putting emphasis on possible solutions, and present evidence of the usefulness of machine learning techniques at the clinical and research levels.

Original languageEnglish (US)
Title of host publicationMachine Learning and Medical Imaging
PublisherElsevier Inc.
Pages299-334
Number of pages36
ISBN (Electronic)9780128041147
ISBN (Print)9780128040768
DOIs
StatePublished - Aug 9 2016
Externally publishedYes

Fingerprint

Learning systems
Medical imaging
Biomarkers
Medicine
Feature extraction
Brain

Keywords

  • Alzheimer's disease
  • Clustering
  • FMRI
  • Genetics
  • Heterogeneity
  • Markov random fields
  • Matrix factorization
  • Multivariate pattern analysis
  • Structural MRI
  • Support vector machines

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Sotiras, A., Gaonkar, B., Eavani, H., Honnorat, N., Varol, E., Dong, A., & Davatzikos, C. (2016). Machine learning as a means toward precision diagnostics and prognostics. In Machine Learning and Medical Imaging (pp. 299-334). Elsevier Inc.. https://doi.org/10.1016/B978-0-12-804076-8.00010-4

Machine learning as a means toward precision diagnostics and prognostics. / Sotiras, A.; Gaonkar, B.; Eavani, H.; Honnorat, N.; Varol, E.; Dong, A.; Davatzikos, C.

Machine Learning and Medical Imaging. Elsevier Inc., 2016. p. 299-334.

Research output: Chapter in Book/Report/Conference proceedingChapter

Sotiras, A, Gaonkar, B, Eavani, H, Honnorat, N, Varol, E, Dong, A & Davatzikos, C 2016, Machine learning as a means toward precision diagnostics and prognostics. in Machine Learning and Medical Imaging. Elsevier Inc., pp. 299-334. https://doi.org/10.1016/B978-0-12-804076-8.00010-4
Sotiras A, Gaonkar B, Eavani H, Honnorat N, Varol E, Dong A et al. Machine learning as a means toward precision diagnostics and prognostics. In Machine Learning and Medical Imaging. Elsevier Inc. 2016. p. 299-334 https://doi.org/10.1016/B978-0-12-804076-8.00010-4
Sotiras, A. ; Gaonkar, B. ; Eavani, H. ; Honnorat, N. ; Varol, E. ; Dong, A. ; Davatzikos, C. / Machine learning as a means toward precision diagnostics and prognostics. Machine Learning and Medical Imaging. Elsevier Inc., 2016. pp. 299-334
@inbook{fec6123ac9ad47c482f66efa7d4699dc,
title = "Machine learning as a means toward precision diagnostics and prognostics",
abstract = "Machine learning plays an essential role in medical imaging. Pattern analysis techniques can identify, and quantify, subtle and spatially complex patterns of disease-induced changes in the brain despite confounding statistical noise and inter-individual variability. This allows the construction of sensitive biomarkers that can identify disease, or risk of developing it, and characterize future clinical progression on an individual patient basis. Thus pattern analysis techniques have become an indispensable part of the growing need for personalized, predictive medicine. However, despite important advances, several challenges remain before they can gain widespread acceptance as tools for precision diagnostics and prognostics in clinical practice. These include: (i) feature extraction and dimensionality reduction; (ii) readily interpreting complex multivariate models; and (iii) elucidating disease heterogeneity. In this chapter, we describe these challenges, putting emphasis on possible solutions, and present evidence of the usefulness of machine learning techniques at the clinical and research levels.",
keywords = "Alzheimer's disease, Clustering, FMRI, Genetics, Heterogeneity, Markov random fields, Matrix factorization, Multivariate pattern analysis, Structural MRI, Support vector machines",
author = "A. Sotiras and B. Gaonkar and H. Eavani and N. Honnorat and E. Varol and A. Dong and C. Davatzikos",
year = "2016",
month = "8",
day = "9",
doi = "10.1016/B978-0-12-804076-8.00010-4",
language = "English (US)",
isbn = "9780128040768",
pages = "299--334",
booktitle = "Machine Learning and Medical Imaging",
publisher = "Elsevier Inc.",

}

TY - CHAP

T1 - Machine learning as a means toward precision diagnostics and prognostics

AU - Sotiras, A.

AU - Gaonkar, B.

AU - Eavani, H.

AU - Honnorat, N.

AU - Varol, E.

AU - Dong, A.

AU - Davatzikos, C.

PY - 2016/8/9

Y1 - 2016/8/9

N2 - Machine learning plays an essential role in medical imaging. Pattern analysis techniques can identify, and quantify, subtle and spatially complex patterns of disease-induced changes in the brain despite confounding statistical noise and inter-individual variability. This allows the construction of sensitive biomarkers that can identify disease, or risk of developing it, and characterize future clinical progression on an individual patient basis. Thus pattern analysis techniques have become an indispensable part of the growing need for personalized, predictive medicine. However, despite important advances, several challenges remain before they can gain widespread acceptance as tools for precision diagnostics and prognostics in clinical practice. These include: (i) feature extraction and dimensionality reduction; (ii) readily interpreting complex multivariate models; and (iii) elucidating disease heterogeneity. In this chapter, we describe these challenges, putting emphasis on possible solutions, and present evidence of the usefulness of machine learning techniques at the clinical and research levels.

AB - Machine learning plays an essential role in medical imaging. Pattern analysis techniques can identify, and quantify, subtle and spatially complex patterns of disease-induced changes in the brain despite confounding statistical noise and inter-individual variability. This allows the construction of sensitive biomarkers that can identify disease, or risk of developing it, and characterize future clinical progression on an individual patient basis. Thus pattern analysis techniques have become an indispensable part of the growing need for personalized, predictive medicine. However, despite important advances, several challenges remain before they can gain widespread acceptance as tools for precision diagnostics and prognostics in clinical practice. These include: (i) feature extraction and dimensionality reduction; (ii) readily interpreting complex multivariate models; and (iii) elucidating disease heterogeneity. In this chapter, we describe these challenges, putting emphasis on possible solutions, and present evidence of the usefulness of machine learning techniques at the clinical and research levels.

KW - Alzheimer's disease

KW - Clustering

KW - FMRI

KW - Genetics

KW - Heterogeneity

KW - Markov random fields

KW - Matrix factorization

KW - Multivariate pattern analysis

KW - Structural MRI

KW - Support vector machines

UR - http://www.scopus.com/inward/record.url?scp=85017470531&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85017470531&partnerID=8YFLogxK

U2 - 10.1016/B978-0-12-804076-8.00010-4

DO - 10.1016/B978-0-12-804076-8.00010-4

M3 - Chapter

AN - SCOPUS:85017470531

SN - 9780128040768

SP - 299

EP - 334

BT - Machine Learning and Medical Imaging

PB - Elsevier Inc.

ER -