Abstract
Background
Deep learning has been employed using brain functional connectome data for evaluating the risk of cognitive deficits in very preterm infants. Although promising, training these deep learning models typically requires a large amount of labeled data, and labeled medical data are often very difficult and expensive to obtain.
Objective
This study aimed to develop a self-training deep neural network (DNN) model for early prediction of cognitive deficits at 2 years of corrected age in very preterm infants (gestational age ≤32 weeks) using both labeled and unlabeled brain functional connectome data.
Materials and methods
We collected brain functional connectome data from 343 very preterm infants at a mean (standard deviation) postmenstrual age of 42.7 (2.5) weeks, among whom 103 children had a cognitive assessment at 2 years (i.e. labeled data), and the remaining 240 children had not received 2-year assessments at the time this study was conducted (i.e. unlabeled data). To develop a self-training DNN model, we built an initial student model using labeled brain functional connectome data. Then, we applied the trained model as a teacher model to generate pseudo-labels for unlabeled brain functional connectome data. Next, we combined labeled and pseudo-labeled data to train a new student model. We iterated this procedure to obtain the best student model for the early prediction task in very preterm infants.
Results
In our cross-validation experiments, the proposed self-training DNN model achieved an accuracy of 71.0%, a specificity of 71.5%, a sensitivity of 70.4% and an area under the curve of 0.75, significantly outperforming transfer learning models through pre-training approaches.
Conclusion
We report the first self-training prognostic study in very preterm infants, efficiently utilizing a small amount of labeled data with a larger share of unlabeled data to aid the model training. The proposed technique is expected to facilitate deep learning with insufficient training data.
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00247-022-05510-8/MediaObjects/247_2022_5510_Fig1_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00247-022-05510-8/MediaObjects/247_2022_5510_Fig2_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00247-022-05510-8/MediaObjects/247_2022_5510_Fig3_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00247-022-05510-8/MediaObjects/247_2022_5510_Fig4_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00247-022-05510-8/MediaObjects/247_2022_5510_Fig5_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00247-022-05510-8/MediaObjects/247_2022_5510_Fig6_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00247-022-05510-8/MediaObjects/247_2022_5510_Fig7_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00247-022-05510-8/MediaObjects/247_2022_5510_Fig8_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00247-022-05510-8/MediaObjects/247_2022_5510_Fig9_HTML.png)
Similar content being viewed by others
References
Ancel P-Y, Goffinet F, Kuhn P et al (2015) Survival and morbidity of preterm children born at 22 through 34 weeks’ gestation in France in 2011: results of the EPIPAGE-2 cohort study. JAMA Pediatr 169:230–238
Vassar R, Schadl K, Cahill-Rowley K et al (2020) Neonatal brain microstructure and machine-learning-based prediction of early language development in children born very preterm. Pediatr Neurol 108:86–92
He L, Parikh NA (2015) Aberrant executive and frontoparietal functional connectivity in very preterm infants with diffuse white matter abnormalities. Pediatr Neurol 53:330–337
Jarjour IT (2015) Neurodevelopmental outcome after extreme prematurity: a review of the literature. Pediatr Neurol 52:143–152
Erdei C, Austin NC, Cherkerzian S et al (2020) Predicting school-aged cognitive impairment in children born very preterm. Pediatrics 145:4
Hack M, Taylor HG, Drotar D et al (2005) Poor predictive validity of the Bayley scales of infant development for cognitive function of extremely low birth weight children at school age. Pediatrics 116:333–341
Ment LR, Vohr B, Allan W et al (2003) Change in cognitive function over time in very low-birth-weight infants. JAMA 289:705–711
Spencer-Smith MM, Spittle AJ, Lee KJ et al (2015) Bayley-III cognitive and language scales in preterm children. Pediatrics 135:1258–1265
Sporns O (2013) The human connectome: origins and challenges. Neuroimage 80:53–61
Glasser MF, Smith SM, Marcus DS et al (2016) The human connectome project’s neuroimaging approach. Nat Neurosci 19:1175–1187
Kim J, Calhoun VD, Shim E et al (2016) Deep neural network with weight sparsity control and pre-training extracts hierarchical features and enhances classification performance: evidence from whole-brain resting-state functional connectivity patterns of schizophrenia. Neuroimage 124:127–146
Kuang D, Guo X, An X et al (2014) Discrimination of ADHD based on fMRI data with deep belief network. In: Huang DS, Han K, Gromiha M (eds) Intelligent computing in bioinformatics. ICIC 2014. Lecture notes in computer science, vol 8590. Springer, Cham, pp 225–232
dos Santos Siqueira A, Biazoli CE Jr, Comfort WE et al (2014) Abnormal functional resting-state networks in ADHD: graph theory and pattern recognition analysis of fMRI data. Biomed Res Int 2014:380531
Heinsfeld AS, Franco AR, Craddock RC et al (2018) Identification of autism spectrum disorder using deep learning and the ABIDE dataset. Neuroimage Clin 17:16–23
He L, Li H, Chen M et al (2021) Deep multimodal learning from MRI and clinical data for early prediction of neurodevelopmental deficits in very preterm infants. Front Neurosci 15:753033
He L, Li H, Wang J et al (2020) A multi-task, multi-stage deep transfer learning model for early prediction of neurodevelopment in very preterm infants. Sci Rep 10:1–13
LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521:436–444
Hjelm RD, Calhoun VD, Salakhutdinov R et al (2014) Restricted Boltzmann machines for neuroimaging: an application in identifying intrinsic networks. Neuroimage 96:245–260
Plis SM, Hjelm DR, Salakhutdinov R et al (2014) Deep learning for neuroimaging: a validation study. Front Neurosci 8:229
Chen M, Li H, Wang J et al (2019) A multichannel deep neural network model analyzing multiscale functional brain connectome data for attention deficit hyperactivity disorder detection. Radiol Artif Intell 2:e190012
Li H, Parikh NA, He L (2018) A novel transfer learning approach to enhance deep neural network classification of brain functional connectomes. Front Neurosci 12:491
He L, Li H, Holland SK et al (2018) Early prediction of cognitive deficits in very preterm infants using functional connectome data in an artificial neural network framework. Neuroimage Clin 18:290–297
Xie Q, Luong M-T, Hovy E, Le QV (2020) Self-training with noisy student improves ImageNet classification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, New York, pp 10687–10698
Pham H, Dai Z, Xie Q et al (2021) Meta pseudo labels. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, New York, pp 11557–11568
Zhai X, Kolesnikov A, Houlsby N et al (2022) Scaling vision transformers. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, New York, pp 12104–12113
He K, Girshick R, Dollár P (2019) Rethinking ImageNet pre-training. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. IEEE, New York, pp 4918–4927
Zoph B, Ghiasi G, Lin T-Y et al (2020) Rethinking pre-training and self-training. In: Advances in Neural Information Processing Systems 33 (NeurIPS 2020), pp 3833–3845
Deng J, Dong W, Socher R et al (2009) ImageNet: a large-scale hierarchical image database. In: 2009 Conference on Computer Vision and Pattern Recognition. IEEE, New York, pp 248–255
Vohr BR, Stephens BE, Higgins RD et al (2012) Are outcomes of extremely preterm infants improving? Impact of Bayley assessment on outcomes. J Pediatr 161:222–228
Aylward GP (2013) Continuing issues with the Bayley-III: where to go from here. J Dev Behav Pediatr 34:697–701
Reuner G, Fields AC, Wittke A et al (2013) Comparison of the developmental tests Bayley-III and Bayley-II in 7-month-old infants born preterm. Eur J Pediatr 172:393–400
Power JD, Barnes KA, Snyder AZ et al (2012) Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuroimage 59:2142–2154
Behzadi Y, Restom K, Liau J, Liu TT (2007) A component based noise correction method (CompCor) for BOLD and perfusion based fMRI. Neuroimage 37:90–101
Shi F, Yap P-T, Wu G et al (2011) Infant brain atlases from neonates to 1- and 2-year-olds. PLoS One 6:e18746
Marrelec G, Krainik A, Duffau H et al (2006) Partial correlation for functional brain interactivity investigation in functional MRI. Neuroimage 32:228–237
Whitfield-Gabrieli S, Nieto-Castanon A (2012) Conn: a functional connectivity toolbox for correlated and anticorrelated brain networks. Brain Connect 2:125–141
Zhao H, Liu F, Li L, Luo C (2018) A novel softplus linear unit for deep convolutional neural networks. Appl Intell 48:1707–1720
Kingma DP, Ba J (2015) Adam: a method for stochastic optimization. Cornell University. https://doi.org/10.48550/arXiv.1412.6980. Accessed 24 July 2022
Di Martino A, Yan C-G, Li Q et al (2014) The Autism Brain Imaging Data Exchange: towards a large-scale evaluation of the intrinsic brain architecture in autism. Mol Psychiatry 19:659–667
Seiffert C, Khoshgoftaar TM, Van Hulse J et al (2009) RUSBoost: a hybrid approach to alleviating class imbalance. IEEE Trans Syst Man Cybern A Syst Hum 40:185–197
Ho TK (1998) Nearest neighbors in random subspaces. In: Joint IAPR International Workshops on Statistical Techniques in Pattern Recognition (SPR) and Structural and Syntactic Pattern Recognition (SSPR), pp 640–648
Freund Y, Schapire RE (1997) A decision-theoretic generalization of on-line learning and an application to boosting. J Comp Syst Sci 55:119–139
Cortes C, Vapnik V (1995) Support-vector networks. Mach Learn 20:273–297
Simonyan K, Vedaldi A, Zisserman A (2013) Deep inside convolutional networks: visualising image classification models and saliency maps. Cornell University. https://doi.org/10.48550/arXiv.1312.6034. Accessed 24 July 2022
Xia M, Wang J, He Y (2013) BrainNet Viewer: a network visualization tool for human brain connectomics. PLoS One 8:e68910
Nair V, Hinton GE (2010) Rectified linear units improve restricted Boltzmann machines. In: Proceedings of the 27th International Conference on Machine Learning (ICML'10). ACM, New York, pp 807–814
Yang Y, Xu Z (2020) Rethinking the value of labels for improving class-imbalanced learning. Adv Neural Inf Process Syst 33:19290–19301
Lafer-Sousa R, Conway BR (2013) Parallel, multi-stage processing of colors, faces and shapes in macaque inferior temporal cortex. Nat Neurosci 16:1870–1878
Hadland KA, Rushworth MFS, Gaffan D, Passingham RE (2003) The effect of cingulate lesions on social behaviour and emotion. Neuropsychologia 41:919–931
Kozlovskiy S, Vartanov A, Pyasik M et al (2013) Anatomical characteristics of cingulate cortex and neuropsychological memory tests performance. Proc Soc Behav Sci 86:128–133
Kozlovskiy SA, Nikonova EY, Pyasik MM et al (2012) The cingulate cortex and human memory processes. Psychol Russia 5:231–243
Drevets WC, Savitz J, Trimble M (2008) The subgenual anterior cingulate cortex in mood disorders. CNS Spectr 13:663–681
Adams R, David AS (2007) Patterns of anterior cingulate activation in schizophrenia: a selective review. Neuropsychiatr Dis Treat 3:87–101
Acknowledgments
We sincerely thank our collaborators from the Cincinnati Infant Neurodevelopment Early Prediction Study (CINEPS) Group: principal investigator Nehal A. Parikh, DO, MS; and collaborators (in alphabetical order) Mekibib Altaye, PhD; Anita Arnsperger, RRT; Traci Beiersdorfer, RN BSN; Kaley Bridgewater, RT(MR) CNMT; Tanya Cahill, MD; Kim Cecil, PhD; Kent Dietrich, RT; Christen Distler, BSN RNC-NIC; Juanita Dudley, RN BSN; Brianne Georg, BS; Cathy Grisby, RN BSN CCRC; Lacey Haas, RT(MR) CNMT; Karen Harpster, PhD, OT/RL; Scott K. Holland, PhD; V.S. Priyanka Illapani, MS; Kristin Kirker, CRC; Julia E. Kline, PhD; Beth M. Kline-Fath, MD; Matt Lanier, RT(MR) RT(R); Stephanie L. Merhar, MD MS; Greg Muthig, BS; Brenda B. Poindexter, MD MS; David Russell, JD; Kari Tepe, BSN RNC-NIC; Leanne Tamm, PhD; Julia Thompson, RN BSN; Jean A. Tkach, PhD; Jinghua Wang, PhD; Brynne Williams, RT(MR) CNMT; Kelsey Wineland, RT(MR) CNMT; Sandra Wuertz, RN BSN CCRP; Donna Wuest, AS; and Weihong Yuan, PhD. We are most grateful to the families that made this study possible.
Funding
This work was supported by the National Institutes of Health (R01-EB029944, R01-EB030582, R01-NS094200 and R01-NS096037); and the Academic and Research Committee (ARC) Awards of Cincinnati Children's Hospital Medical Center. The funders played no role in the design, analysis or presentation of the findings.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflicts of interest
None
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Ali, R., Li, H., Dillman, J.R. et al. A self-training deep neural network for early prediction of cognitive deficits in very preterm infants using brain functional connectome data. Pediatr Radiol 52, 2227–2240 (2022). https://doi.org/10.1007/s00247-022-05510-8
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00247-022-05510-8