scholarly journals Is whole-brain functional connectivity a neuromarker of sustained attention? Comment on Rosenberg & al. (2016)

2017 ◽  
Author(s):  
Simon Bang Kristensen ◽  
Kristian Sandberg

AbstractIdentification of neuromarkers accurately predicting cognitive characteristics from a single, standardised MRI scan could be tremendously useful in basic psychology and clinical practise. In a recent article, Rosenberg et al. (Nat Neurosci, 19, 165-171, 2016) argue that whole-brain functional network strength is a broadly applicable neuromarker of sustained attention. They claim that this marker accurately predicts performance from task related as well as resting state activity. Here, we discuss the applicability and generalizability in the context of three methodological concerns: Simulations show that the statistical methods for the 1) initial validation analyses as well as 2) internal validation using leave-one-out cross validation, are biased towards significance; 3) simple and complex models are compared suboptimally. Overall, we find that the article of Rosenberg et al. provides sufficient proof that network strength is associated with attentional capacity that it is not possible to say to which extent, and for this reason we argue that it cannot be concluded that the network is a broadly applicable neuromarker.

2015 ◽  
Vol 19 (1) ◽  
pp. 165-171 ◽  
Author(s):  
Monica D Rosenberg ◽  
Emily S Finn ◽  
Dustin Scheinost ◽  
Xenophon Papademetris ◽  
Xilin Shen ◽  
...  

2018 ◽  
Vol 14 (1) ◽  
pp. 100-109 ◽  
Author(s):  
Jinliang Zhang ◽  
Gaoyan Zhang ◽  
Xianglin Li ◽  
Peiyuan Wang ◽  
Bin Wang ◽  
...  

2017 ◽  
Vol 6 (2) ◽  
pp. 243-265 ◽  
Author(s):  
Eric A. Youngstrom ◽  
Tate F. Halverson ◽  
Jennifer K. Youngstrom ◽  
Oliver Lindhiem ◽  
Robert L. Findling

Reliability of clinical diagnoses is often low. There are many algorithms that could improve diagnostic accuracy, and statistical learning is becoming popular. Using pediatric bipolar disorder as a clinically challenging example, we evaluated a series of increasingly complex models ranging from simple screening to a supervised LASSO (least absolute shrinkage and selection operation) regression in a large ( N = 550) academic clinic sample. We then externally validated models in a community clinic ( N = 511) with the same candidate predictors and semistructured interview diagnoses, providing high methodological consistency; the clinics also had substantially different demography and referral patterns. Models performed well according to internal validation metrics. Complex models degraded rapidly when externally validated. Naive Bayesian and logistic models concentrating on predictors identified in prior meta-analyses tied or bettered LASSO models when externally validated. Implementing these methods would improve clinical diagnostic performance. Statistical learning research should continue to invest in high-quality indicators and diagnoses to supervise model training.


Sign in / Sign up

Export Citation Format

Share Document