TY - JOUR
T1 - The Heterogeneity Problem
T2 - Approaches to Identify Psychiatric Subtypes
AU - Feczko, Eric
AU - Miranda-Dominguez, Oscar
AU - Marr, Mollie
AU - Graham, Alice M.
AU - Nigg, Joel T.
AU - Fair, Damien A.
N1 - Funding Information:
This research was supported by DeStefano Family Foundation , United States of America; the National Library of Medicine ( T15 LM007088 ), United States of America; and the National Institute of Mental Health ( R01 MH096773 , R00 MH091238 , R01 MH096773-03S1 , R01 MH 096773–05 , R01 MH086654 , R01 MH086654 , R01 MH59107 ), United States of America.
Publisher Copyright:
© 2019
PY - 2019/7
Y1 - 2019/7
N2 - The imprecise nature of psychiatric nosology restricts progress towards characterizing and treating mental health disorders. One issue is the ‘heterogeneity problem’: different causal mechanisms may relate to the same disorder, and multiple outcomes of interest can occur within one individual. Our review tackles this heterogeneity problem, providing considerations, concepts, and approaches for investigators examining human cognition and mental health. We highlight the difficulty of pure dimensional approaches due to ‘the curse of dimensionality’. Computationally, we consider supervised and unsupervised statistical approaches to identify putative subtypes within a population. However, we emphasize that subtype identification should be linked to a particular outcome or question. We conclude with novel hybrid approaches that can identify subtypes tied to outcomes, and may help advance precision diagnostic and treatment tools.
AB - The imprecise nature of psychiatric nosology restricts progress towards characterizing and treating mental health disorders. One issue is the ‘heterogeneity problem’: different causal mechanisms may relate to the same disorder, and multiple outcomes of interest can occur within one individual. Our review tackles this heterogeneity problem, providing considerations, concepts, and approaches for investigators examining human cognition and mental health. We highlight the difficulty of pure dimensional approaches due to ‘the curse of dimensionality’. Computationally, we consider supervised and unsupervised statistical approaches to identify putative subtypes within a population. However, we emphasize that subtype identification should be linked to a particular outcome or question. We conclude with novel hybrid approaches that can identify subtypes tied to outcomes, and may help advance precision diagnostic and treatment tools.
KW - functional random forest
KW - heterogeneity
KW - machine learning
KW - mental health
KW - surrogate variable analysis
UR - http://www.scopus.com/inward/record.url?scp=85066289449&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85066289449&partnerID=8YFLogxK
U2 - 10.1016/j.tics.2019.03.009
DO - 10.1016/j.tics.2019.03.009
M3 - Review article
C2 - 31153774
AN - SCOPUS:85066289449
SN - 1364-6613
VL - 23
SP - 584
EP - 601
JO - Trends in Cognitive Sciences
JF - Trends in Cognitive Sciences
IS - 7
ER -