| DLUHOŠ, Petr, Daniel SCHWARZ, Wiepke CAHN, Neeltje van HAREN, René KAHN, Filip ŠPANIEL, Jiří HORÁČEK, Tomáš KAŠPÁREK and Hugo SCHNACK. Multi-center machine learning in imaging psychiatry: A meta-model approach. NeuroImage. San Diego: Academic Press Inc Elsevier Science, 2017, vol. 155, 15 July 2017, p. 10-24. ISSN 1053-8119. Available from: https://dx.doi.org/10.1016/j.neuroimage.2017.03.027. |
Other formats:
BibTeX
LaTeX
RIS
@article{1386486,
author = {Dluhoš, Petr and Schwarz, Daniel and Cahn, Wiepke and Haren, Neeltje van and Kahn, René and Španiel, Filip and Horáček, Jiří and Kašpárek, Tomáš and Schnack, Hugo},
article_location = {San Diego},
article_number = {15 July 2017},
doi = {http://dx.doi.org/10.1016/j.neuroimage.2017.03.027},
keywords = {classification; combining models; first-episode schizophrenia; Machine learning; multi-center; prediction; support vector machines (SVM)},
language = {eng},
issn = {1053-8119},
journal = {NeuroImage},
title = {Multi-center machine learning in imaging psychiatry: A meta-model approach},
volume = {155},
year = {2017}
}
TY - JOUR
ID - 1386486
AU - Dluhoš, Petr - Schwarz, Daniel - Cahn, Wiepke - Haren, Neeltje van - Kahn, René - Španiel, Filip - Horáček, Jiří - Kašpárek, Tomáš - Schnack, Hugo
PY - 2017
TI - Multi-center machine learning in imaging psychiatry: A meta-model approach
JF - NeuroImage
VL - 155
IS - 15 July 2017
SP - 10-24
EP - 10-24
PB - Academic Press Inc Elsevier Science
SN - 10538119
KW - classification
KW - combining models
KW - first-episode schizophrenia
KW - Machine learning
KW - multi-center
KW - prediction
KW - support vector machines (SVM)
N2 - One of the biggest problems in automated diagnosis of psychiatric disorders from medical images is the lack of sufficiently large samples for training. Sample size is especially important in the case of highly heterogeneous disorders such as schizophrenia, where machine learning models built on relatively low numbers of subjects may suffer from poor generalizability. Via multicenter studies and consortium initiatives researchers have tried to solve this problem by combining data sets from multiple sites. The necessary sharing of (raw) data is, however, often hindered by legal and ethical issues. Moreover, in the case of very large samples, the computational complexity might become too large. The solution to this problem could be distributed learning. In this paper we investigated the possibility to create a meta-model by combining support vector machines (SVM) classifiers trained on the local datasets, without the need for sharing medical images or any other personal data. Validation was done in a 4-center setup comprising of 480 first-episode schizophrenia patients and healthy controls in total. We built SVM models to separate patients from controls based on three different kinds of imaging features derived from structural MRI scans, and compared models built on the joint multicenter data to the meta-models. The results showed that the combined meta-model had high similarity to the model built on all data pooled together and comparable classification performance on all three imaging features. Both similarity and performance was superior to that of the local models. We conclude that combining models is thus a viable alternative that facilitates data sharing and creating bigger and more informative models.
ER -
DLUHOŠ, Petr, Daniel SCHWARZ, Wiepke CAHN, Neeltje van HAREN, René KAHN, Filip ŠPANIEL, Jiří HORÁČEK, Tomáš KAŠPÁREK and Hugo SCHNACK. Multi-center machine learning in imaging psychiatry: A meta-model approach. \textit{NeuroImage}. San Diego: Academic Press Inc Elsevier Science, 2017, vol.~155, 15 July 2017, p.~10-24. ISSN~1053-8119. Available from: https://dx.doi.org/10.1016/j.neuroimage.2017.03.027.
|
