CoChat

A computational framework for the detection of subcortical brain dysmaturation in neonatal MRI using 3D Convolutional Neural Networks

Open access
Ceschin, Rafael, Zahner, Alexandria, Reynolds, William, Gaesser, Jenna, Zuccoli, Giulio, Lo, Cecilia W., Gopalakrishnan, Vanathi, Panigrahy, Ashok · 2018 · NeuroImage
Cited by 45

The DOI matches its Crossref record; 12 of the 12 references it cites were found in a registry or an index.

Abstract

Abstract Deep neural networks are increasingly being used in both supervised learning for classification tasks and unsupervised learning to derive complex patterns from the input data. However, the successful implementation of deep neural networks using neuroimaging datasets requires adequate sample size for training and well‐defined signal intensity based structural differentiation. There is a lack of effective automated diagnostic tools for the reliable detection of brain dysmaturation in the neonatal period, related to small sample size and complex undifferentiated brain structures, despite both translational research and clinical importance. Volumetric information alone is insufficient for diagnosis. In this study, we developed a computational framework for the automated classification of brain dysmaturation from neonatal MRI, by combining a specific deep neural network implementation with neonatal structural brain segmentation as a method for both clinical pattern recognition and data‐driven inference into the underlying structural morphology. We implemented three‐dimensional convolution neural networks (3D‐CNNs) to specifically classify dysplastic cerebelli, a subset of surface‐based subcortical brain dysmaturation, in term infants born with congenital heart disease. We obtained a 0.985 ± 0. 0241‐classification accuracy of subtle cerebellar dysplasia in CHD using 10‐fold cross‐validation. Furthermore, the hidden layer activations and class activation maps depicted regional vulnerability of the superior surface of the cerebellum, (composed of mostly the posterior lobe and the midline vermis), in regards to differentiating the dysplastic process from normal tissue. The posterior lobe and the midline vermis provide regional differentiation that is relevant to not only to the clinical diagnosis of cerebellar dysplasia, but also genetic mechanisms and neurodevelopmental outcome correlates. These findings not only contribute to the detection and classification of a subset of neonatal brain dysmaturation, but also provide insight to the pathogenesis of cerebellar dysplasia in CHD. In addition, this is one of the first examples of the application of deep learning to a neuroimaging dataset, in which the hidden layer activation revealed diagnostically and biologically relevant features about the clinical pathogenesis. The code developed for this project is open source, published under the BSD License, and designed to be generalizable to applications both within and beyond neonatal brain imaging. HighlightsA framework for classification of brain dysmaturation from neonatal MRI is introduced.Combine deep learning and structural brain segmentation.Provides both clinical pattern recognition and data‐driven inference.Revealed diagnostic and biologic insight about clinical pathogenesis.The code developed is open source, published under the BSD License.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1016/j.neuroimage.2018.05.049

Cites

12 of 12 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 12 references

Research works printed but not linked: 42

  1. Aguilar, 2012, Analysis of human samples reveals impaired SHH-dependent cerebellar development in Joubert syndrome/Meckel syndrome, Proc. Natl. Acad. Sci. Unit. States Am., 109, 16951, 10.1073/pnas.1201408109 · DOI
  2. Allin, 2005, Vermis and lateral lobes of the cerebellum in adolescents born very preterm, Neuroreport, 16, 1821, 10.1097/01.wnr.0000185014.36939.84 · DOI
  3. Avants, 2009, Advanced normalization tools (ANTS), Insight J, 1
  4. Bellinger, 2010, Neuropsychological, psychosocial, and quality-of-life outcomes in children and adolescents with congenital heart disease, Prog. Pediatr. Cardiol., 29, 87, 10.1016/j.ppedcard.2010.06.007 · DOI
  5. Bolduc, 2012, Regional cerebellar volumes predict functional outcome in children with cerebellar malformations, Cerebellum, 11, 531, 10.1007/s12311-011-0312-z · DOI · Paper page
  6. Bosemani, 2016, Cerebellar disruptions and neurodevelopmental disabilities, Semin. Fetal Neonatal Med., 1
  7. Brossard-Racine, 2015, Developmental cerebellar cognitive affective syndrome in ex-preterm survivors following cerebellar injury, Cerebellum, 14, 151, 10.1007/s12311-014-0597-9 · DOI
  8. Doherty, 2013, Midbrain-hindbrain malformations: advances in clinical diagnosis, imaging, and genetics, Lancet Neurol., 12, 381, 10.1016/S1474-4422(13)70024-3 · DOI
  9. D'Angelo, 2012, Seeking a unified framework for cerebellar function and dysfunction: from circuit operations to cognition, Front. Neural Circ., 6, 116
  10. Erhan, 2009, Visualizing higher-layer features of a deep network, Bernoulli, 1
  11. Gorgolewski, 2011, Nipype: a flexible, lightweight and extensible neuroimaging data processing framework in python, Front. Neuroinf., 5, 10.3389/fninf.2011.00013 · DOI · Paper page
  12. Gousias, 2012, Magnetic resonance imaging of the newborn brain: manual segmentation of labelled atlases in term-born and preterm infants, Neuroimage, 62, 1499, 10.1016/j.neuroimage.2012.05.083 · DOI · Paper page
  13. Gousias, 2013, Magnetic resonance imaging of the newborn brain: automatic segmentation of brain images into 50 anatomical regions, vol. 8
  14. Gupta A., Ayhan M.S., Maida A.S. Natural Image Bases to Represent Neuroimaging Data. http://proceedings.mlr.press/v28/gupta13b.pdf. Accessed 4 October 2017.
  15. Haldipur, 2017, Phenotypic outcomes in Mouse and Human Foxc1 dependent Dandy-Walker cerebellar malformation suggest shared mechanisms, Elife, 6, 1, 10.7554/eLife.20898 · DOI
  16. Hua, 2008, Tensor-based morphometry as a neuroimaging biomarker for Alzheimer's disease: an MRI study of 676 AD, MCI, and normal subjects, Neuroimage, 43, 458, 10.1016/j.neuroimage.2008.07.013 · DOI · Paper page
  17. Jaware, 2016, Multi-kernel support vector machine and Levenberg-Marquardt classification approach for neonatal brain MR images, vol. 8, 1
  18. Jenkinson, 2012, FSL, Neuroimage, 62, 782, 10.1016/j.neuroimage.2011.09.015 · DOI · Paper page
  19. Li, 2015, Global genetic analysis in mice unveils central role for cilia in congenital heart disease, Nature, 521, 520, 10.1038/nature14269 · DOI · Paper page
  20. Limperopoulos, 2010, Cerebellar injury in the premature infant is associated with impaired growth of specific cerebral regions, Pediatr. Res., 68, 145, 10.1203/PDR.0b013e3181e1d032 · DOI
  21. Limperopoulos, 2014, Injury to the premature cerebellum: outcome is related to remote cortical development, Cerebr. Cortex, 24, 728, 10.1093/cercor/bhs354 · DOI · Paper page
  22. Liu, 2017, The complex genetics of hypoplastic left heart syndrome, Nat. Genet., 49, 1152, 10.1038/ng.3870 · DOI · Paper page
  23. Mueller, 2005, Ways toward an early diagnosis in Alzheimer's disease: the Alzheimer's disease neuroimaging initiative (ADNI), Alzheimer's Dementia, 1, 55, 10.1016/j.jalz.2005.06.003 · DOI · Paper page
  24. Oegema, 2015, Recognizable cerebellar dysplasia associated with mutations in multiple tubulin genes, Hum. Mol. Genet., 24, 5313, 10.1093/hmg/ddv250 · DOI
  25. Ortinau, 2012, Congenital heart disease affects cerebral size but not brain growth, Pediatr. Cardiol., 33, 1138, 10.1007/s00246-012-0269-9 · DOI · Paper page
  26. Panigrahy, 2016, Brain dysplasia associated with ciliary dysfunction in infants with congenital heart disease, J. Pediatr., 178, 10.1016/j.jpeds.2016.07.041 · DOI
  27. Poretti, 2016, Huisman TAGM. Pre- and postnatal neuroimaging of congenital cerebellar abnormalities, Cerebellum, 15, 5, 10.1007/s12311-015-0699-z · DOI
  28. Quinn, 2015, Automated identification of abnormal respiratory ciliary motion in nasal biopsies, Sci. Transl. Med., 7, 299ra124, 10.1126/scitranslmed.aaa1233 · DOI
  29. Schmahmann, 1998, The cerebellar cognitive affective syndrome, Brain, 121, 561, 10.1093/brain/121.4.561 · DOI · Paper page
  30. Song, 2007, Clinical neonatal brain MRI segmentation and intensity-based markov priors, Computer, 1
  31. Srhoj-Egekher, 2012, Automatic segmentation of neonatal brain MRI using atlas based segmentation and machine learning approach, MICCAI Gd Chall Neonatal Brain Segmentation
  32. Stoodley, 2016, Structure–function relationships in the developing cerebellum: evidence from early-life cerebellar injury and neurodevelopmental disorders, Semin. Fetal Neonatal Med., 21, 1, 10.1016/j.siny.2016.04.010 · DOI · Paper page
  33. Tam, 2013, Potential mechanisms of cerebellar hypoplasia in prematurity, Neuroradiology, 55, 41, 10.1007/s00234-013-1230-1 · DOI
  34. Tiemeier, 2010, Cerebellum development during childhood and adolescence: a longitudinal morphometric MRI study, Neuroimage, 49, 63, 10.1016/j.neuroimage.2009.08.016 · DOI · Paper page
  35. Varoquaux, 2017, Assessing and tuning brain decoders: cross-validation, caveats, and guidelines, Neuroimage, 145, 166, 10.1016/j.neuroimage.2016.10.038 · DOI · Paper page
  36. Vieira, 2017, Using deep learning to investigate the neuroimaging correlates of psychiatric and neurological disorders: methods and applications, Neurosci. Biobehav. Rev., 74, 58, 10.1016/j.neubiorev.2017.01.002 · DOI · Paper page
  37. Vovk, 2007, A review of methods for correction of intensity inhomogeneity in MRI, IEEE Trans. Med. Imag., 26, 405, 10.1109/TMI.2006.891486 · DOI · Paper page
  38. Wegiel, 2010, The neuropathology of autism: defects of neurogenesis and neuronal migration, and dysplastic changes, Acta Neuropathol., 119, 755, 10.1007/s00401-010-0655-4 · DOI · Paper page
  39. Wong, 2017, Synchronous aberrant cerebellar and opercular development in fetuses and neonates with congenital heart disease: correlation with early communicative neurodevelopmental outcomes, initial experience, Am. J. Perinatol. Rep., 7, 17, 10.1055/s-0036-1597934 · DOI
  40. Zeiler, 2014, Visualizing and understanding convolutional networks, Comput Vision–ECCV, 8689, 818
  41. Zhang, 2001, Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm, IEEE Trans. Med. Imag., 20, 45, 10.1109/42.906424 · DOI · Paper page
  42. Zou, 2004, Statistical validation of image segmentation quality based on a spatial overlap index, Acad. Radiol., 11, 178, 10.1016/S1076-6332(03)00671-8 · DOI · Paper page

Cited by

See all 45

Metadata

Authors
Ceschin, R., Zahner, A., Reynolds, W., Gaesser, J., Zuccoli, G., Lo, C. W., Gopalakrishnan, V., & Panigrahy, A.
Venue
NeuroImage
Publisher
Elsevier BV
Published
2018-09
Volume
178
Pages
183-197
ISSN
1053-8119 (ISSN-L)
OpenAlex
W2804692111
Cited by
45
References
12
FWCI
5.15
Type
journal article
Language
en
Open access
bronze
Funders
  • National Institute of Neurological Disorders and Stroke (K23-063371)
  • U.S. National Library of Medicine (5T15LM007059-27)
  • Pennsylvania Department of Health
  • National Heart, Lung, and Blood Institute (R01 HL128818-03)
  • U.S. Department of Defense (W81XWH-16-1-0613)

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: A computational framework for the detection of subcortical brain dysmaturation in neonatal MRI using 3D Convolutional Neural Networks. https://app.cochat.ai/papers/10.1016/j.neuroimage.2018.05.049
The paper
Ceschin, R., Zahner, A., Reynolds, W., Gaesser, J., Zuccoli, G., Lo, C. W., Gopalakrishnan, V., & Panigrahy, A. (2018). A computational framework for the detection of subcortical brain dysmaturation in neonatal MRI using 3D Convolutional Neural Networks. NeuroImage. https://doi.org/10.1016/j.neuroimage.2018.05.049

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.