Convolutional neural network-based classification of craniosynostosis and suture lines from multi-view cranial X-rays

Abstract Early and precise diagnosis of craniosynostosis (CSO), which involves premature fusion of cranial sutures in infants, is crucial for effective treatment. Although computed topography offers detailed imaging, its high radiation poses risks, especially to children. Therefore, we propose a dee...

Full description

Saved in:
Bibliographic Details
Main Authors: Seung Min Kim, Ji Seung Yang, Jae Woong Han, Hyung Il Koo, Tae Hoon Roh, Soo Han Yoon
Format: Article
Language:English
Published: Nature Portfolio 2024-11-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-024-77550-z
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846172037141233664
author Seung Min Kim
Ji Seung Yang
Jae Woong Han
Hyung Il Koo
Tae Hoon Roh
Soo Han Yoon
author_facet Seung Min Kim
Ji Seung Yang
Jae Woong Han
Hyung Il Koo
Tae Hoon Roh
Soo Han Yoon
author_sort Seung Min Kim
collection DOAJ
description Abstract Early and precise diagnosis of craniosynostosis (CSO), which involves premature fusion of cranial sutures in infants, is crucial for effective treatment. Although computed topography offers detailed imaging, its high radiation poses risks, especially to children. Therefore, we propose a deep-learning model for CSO and suture-line classification using 2D cranial X-rays that minimises radiation-exposure risks and offers reliable diagnoses. We used data comprising 1,047 normal and 277 CSO cases from 2006 to 2023. Our approach integrates X-ray-marker removal, head-pose standardisation, skull-cropping, and fine-tuning modules for CSO and suture-line classification using convolution neural networks (CNNs). It enhances the diagnostic accuracy and efficiency of identifying CSO from X-ray images, offering a promising alternative to traditional methods. Four CNN backbones exhibited robust performance, with F1-scores exceeding 0.96 and sensitivity and specificity exceeding 0.9, proving the potential for clinical applications. Additionally, preprocessing strategies further enhanced the accuracy, demonstrating the highest F1-scores, precision, and specificity. A qualitative analysis using gradient-weighted class activation mapping illustrated the focal points of the models. Furthermore, the suture-line classification model distinguishes five suture lines with an accuracy of > 0.9. Thus, the proposed approach can significantly reduce the time and labour required for CSO diagnosis, streamlining its management in clinical settings.
format Article
id doaj-art-728450b777d7425d97d917ac7e124d5d
institution Kabale University
issn 2045-2322
language English
publishDate 2024-11-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-728450b777d7425d97d917ac7e124d5d2024-11-10T12:21:07ZengNature PortfolioScientific Reports2045-23222024-11-0114111110.1038/s41598-024-77550-zConvolutional neural network-based classification of craniosynostosis and suture lines from multi-view cranial X-raysSeung Min Kim0Ji Seung Yang1Jae Woong Han2Hyung Il Koo3Tae Hoon Roh4Soo Han Yoon5Department of Artificial Intelligence, Ajou UniversityDepartment of Industrial Engineering, Ajou UniversityDepartment of Artificial Intelligence, Ajou UniversityDepartment of Electrical Computer Engineering, Ajou UniversityDepartment of Neurosurgery, Ajou University School of MedicineDepartment of Neurosurgery, Ajou University School of MedicineAbstract Early and precise diagnosis of craniosynostosis (CSO), which involves premature fusion of cranial sutures in infants, is crucial for effective treatment. Although computed topography offers detailed imaging, its high radiation poses risks, especially to children. Therefore, we propose a deep-learning model for CSO and suture-line classification using 2D cranial X-rays that minimises radiation-exposure risks and offers reliable diagnoses. We used data comprising 1,047 normal and 277 CSO cases from 2006 to 2023. Our approach integrates X-ray-marker removal, head-pose standardisation, skull-cropping, and fine-tuning modules for CSO and suture-line classification using convolution neural networks (CNNs). It enhances the diagnostic accuracy and efficiency of identifying CSO from X-ray images, offering a promising alternative to traditional methods. Four CNN backbones exhibited robust performance, with F1-scores exceeding 0.96 and sensitivity and specificity exceeding 0.9, proving the potential for clinical applications. Additionally, preprocessing strategies further enhanced the accuracy, demonstrating the highest F1-scores, precision, and specificity. A qualitative analysis using gradient-weighted class activation mapping illustrated the focal points of the models. Furthermore, the suture-line classification model distinguishes five suture lines with an accuracy of > 0.9. Thus, the proposed approach can significantly reduce the time and labour required for CSO diagnosis, streamlining its management in clinical settings.https://doi.org/10.1038/s41598-024-77550-zCraniosynostosisSuture lineTransfer learningDeep learningConvolutional neural networkSkull X-ray
spellingShingle Seung Min Kim
Ji Seung Yang
Jae Woong Han
Hyung Il Koo
Tae Hoon Roh
Soo Han Yoon
Convolutional neural network-based classification of craniosynostosis and suture lines from multi-view cranial X-rays
Scientific Reports
Craniosynostosis
Suture line
Transfer learning
Deep learning
Convolutional neural network
Skull X-ray
title Convolutional neural network-based classification of craniosynostosis and suture lines from multi-view cranial X-rays
title_full Convolutional neural network-based classification of craniosynostosis and suture lines from multi-view cranial X-rays
title_fullStr Convolutional neural network-based classification of craniosynostosis and suture lines from multi-view cranial X-rays
title_full_unstemmed Convolutional neural network-based classification of craniosynostosis and suture lines from multi-view cranial X-rays
title_short Convolutional neural network-based classification of craniosynostosis and suture lines from multi-view cranial X-rays
title_sort convolutional neural network based classification of craniosynostosis and suture lines from multi view cranial x rays
topic Craniosynostosis
Suture line
Transfer learning
Deep learning
Convolutional neural network
Skull X-ray
url https://doi.org/10.1038/s41598-024-77550-z
work_keys_str_mv AT seungminkim convolutionalneuralnetworkbasedclassificationofcraniosynostosisandsuturelinesfrommultiviewcranialxrays
AT jiseungyang convolutionalneuralnetworkbasedclassificationofcraniosynostosisandsuturelinesfrommultiviewcranialxrays
AT jaewoonghan convolutionalneuralnetworkbasedclassificationofcraniosynostosisandsuturelinesfrommultiviewcranialxrays
AT hyungilkoo convolutionalneuralnetworkbasedclassificationofcraniosynostosisandsuturelinesfrommultiviewcranialxrays
AT taehoonroh convolutionalneuralnetworkbasedclassificationofcraniosynostosisandsuturelinesfrommultiviewcranialxrays
AT soohanyoon convolutionalneuralnetworkbasedclassificationofcraniosynostosisandsuturelinesfrommultiviewcranialxrays