TY - CHAP
T1 - Detection of Injury and Automated Triage of Preterm Neonatal MRI Using Patch-Based Gaussian Processes
AU - Macleod, Russell
AU - Counsell, Serena
AU - Carmichael, David
AU - Dimitrova, Ralica
AU - Pietsch, Maximilian
AU - Edwards, A. David
AU - Rutherford, Mary Ann
AU - O’Muircheartaigh, Jonathan
N1 - Funding Information:
This work was supported by the Wellcome/EPSRC Centre for Medical Engineering at Kings College London (WT 203148/Z/16/Z), the NIHR Clinical Research Facility (CRF) at Guy?s and St Thomas? and by the National Institute for Health Research Biomedical Research Centres based at Guy?s and St Thomas? NHS Foundation Trust, and South London, Maudsley NHS Foundation Trust. J.O. is supported by a Sir Henry Dale Fellowship jointly funded by the Wellcome Trust and the Royal Society (grant 206675/Z/17/Z). J.O. received support from the Medical Research Council Centre for Neurodevelopmental Disorders, King?s College London (grant MR/N026063/1). The project includes data from a programme of research funded by the NIHR Programme Grants for Applied Research Programme (RP-PG-0707-10154.)
Publisher Copyright:
© 2021, Springer Nature Switzerland AG.
PY - 2021
Y1 - 2021
N2 - Automatic detection or highlighting of neonatal brain injury could be a valuable adjunct to radiological interpretation. Here we propose a normative modeling-based detection method for preterm neonatal neuroimaging using gaussian processes (GPs). These GPs incorporates local image intensity information from image patches and demographics such as age. Z-score images can then be created from the scaled difference between the model predictions and a neonate’s T1 and T2 weighted MRI. To test the use of these GP Z-scores as a form of automated triage, we trained a logistic regression classifier to separate normal and abnormal images. We used 133 preterm neonatal images with normal-reported MRI to train a GP model and optimized lesion detection performance on 36 preterm neonatal images with manually annotated lesion masks. The automated triage model was trained on 100 preterm neonates with normal reported MRI and 109 preterm neonates with MRI detectable lesions. It was tested on the same 36 manually annotated abnormal MRI preterm neonates and 33 normal-reported preterm neonates. Using a patch diameter of 7 voxels and integrating both T1w and T2w Z-score images provided our highest performing GP model for within image lesion detection, achieving an AUC of 0.961. By combining the output probabilities of a T1w and a T2w Z-score histogram classifiers allows for the correctly identification of 32/36 abnormal and 28/33 normal images. These results indicate patch-based normative model can accurately detect lesions in a highly interpretable fashion in preterm neonates with abnormal MRI. Using outputs from these predictions, the classifier is effective at separating abnormal and normal images.
AB - Automatic detection or highlighting of neonatal brain injury could be a valuable adjunct to radiological interpretation. Here we propose a normative modeling-based detection method for preterm neonatal neuroimaging using gaussian processes (GPs). These GPs incorporates local image intensity information from image patches and demographics such as age. Z-score images can then be created from the scaled difference between the model predictions and a neonate’s T1 and T2 weighted MRI. To test the use of these GP Z-scores as a form of automated triage, we trained a logistic regression classifier to separate normal and abnormal images. We used 133 preterm neonatal images with normal-reported MRI to train a GP model and optimized lesion detection performance on 36 preterm neonatal images with manually annotated lesion masks. The automated triage model was trained on 100 preterm neonates with normal reported MRI and 109 preterm neonates with MRI detectable lesions. It was tested on the same 36 manually annotated abnormal MRI preterm neonates and 33 normal-reported preterm neonates. Using a patch diameter of 7 voxels and integrating both T1w and T2w Z-score images provided our highest performing GP model for within image lesion detection, achieving an AUC of 0.961. By combining the output probabilities of a T1w and a T2w Z-score histogram classifiers allows for the correctly identification of 32/36 abnormal and 28/33 normal images. These results indicate patch-based normative model can accurately detect lesions in a highly interpretable fashion in preterm neonates with abnormal MRI. Using outputs from these predictions, the classifier is effective at separating abnormal and normal images.
KW - Abnormality detection
KW - Gaussian processes
KW - Neonatal imaging
KW - Normative modeling
UR - http://www.scopus.com/inward/record.url?scp=85117116138&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-87735-4_22
DO - 10.1007/978-3-030-87735-4_22
M3 - Conference paper
AN - SCOPUS:85117116138
SN - 9783030877347
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 231
EP - 241
BT - Uncertainty for Safe Utilization of Machine Learning in Medical Imaging, and Perinatal Imaging, Placental and Preterm Image Analysis - 3rd International Workshop, UNSURE 2021, and 6th International Workshop, PIPPI 2021, Held in Conjunction with MICCAI 2021, Proceedings
A2 - Sudre, Carole H.
A2 - Licandro, Roxane
A2 - Baumgartner, Christian
A2 - Melbourne, Andrew
A2 - Dalca, Adrian
A2 - Hutter, Jana
A2 - Tanno, Ryutaro
A2 - Abaci Turk, Esra
A2 - Van Leemput, Koen
A2 - Torrents Barrena, Jordina
A2 - Wells, William M.
A2 - Macgowan, Christopher
PB - Springer Science and Business Media Deutschland GmbH
T2 - 3rd International Workshop on Uncertainty for Safe Utilization of Machine Learning in Medical Imaging, UNSURE 2021, and the 6th International Workshop on Perinatal, Preterm and Paediatric Image Analysis, PIPPI 2021, held in conjunction with the 24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021
Y2 - 1 October 2021 through 1 October 2021
ER -