Fetüs beyin dokularının 3B U-Net ile segmentasyonu ve gestasyonel yaşın segmentasyon performansına etkisi

Gebelik esnasında fetüs beyninde, çevresel veya genetik etmenlerden kaynaklanan bozuklukların ilerleyen yaşlarda otizm, hiperaktivite ve bipolar bozukluklar olarak ortaya çıktığı düşünülmektedir. Fetüs beyin dokularının yapısal analizi, dokuların büyüklükleri ve şekilleri hakkında bilgi sağlayarak bu hastalıkların etiyolojisinin araştırılmasına yardımcı olmaktadır. MR (manyetik rezonans) ve ultrason fetüs beyin dokusu analizi için en sık kullanılan iki görüntüleme tekniğidir. Özellikle MR görüntülemede, fetüs beyin dokuları net bir şekilde görülebilmesine rağmen bu görüntülerin yapısal analizi zaman alıcı bir iştir. Hızla gelişen ve değişen fetüs beyin dokuları, haftalık veya aylık periyotlarla, elle veya yarı-otomatik şekilde gerçekleştirilen MR analizinin yapılmasını zorlaştırmaktadır. Bu araştırmada, 3B U-Net ile MR görüntüleri üzerinde yedi farklı fetüs beyin dokusunun tam otomatik segmentasyonu gerçekleştirilmiştir. FeTA2021 veri seti, erken, orta ve geç gestasyonel yaş gruplarına bölünerek her bir yaş grubu için farklı 3B U-Net modelleri eğitilmiş ve segmentasyon performansı analiz edilmiştir. Gestasyonel yaş gruplarına göre sırasıyla ortalama 0.83, 0.91 ve 0.92 Dice skoruna ulaşılmıştır.

Anahtar Kelimeler:

fetüs beyin dokusu, MR görüntüleme, segmentasyon, U-Net

Segmentation of fetal brain tissues using 3D U-Net and the effect of gestational age on segmentation performance

Environmental and genetic factors are thought to cause diseases in the fetal brain that may later manifest as autism, hyperactivity, and bipolar disorder. Structural analysis of fetal brain tissue helps to explore the etiology of these diseases by providing information about the size and shape of the tissues. MR (magnetic resonance) imaging and ultrasound are the two imaging modalities most used to analyze fetal brain tissue. Especially in MR imaging, brain tissues can be seen clearly. However, structural analysis of MR images is a time-consuming process. Because fetal brain tissues develop and change rapidly, it is difficult to perform manual or semi-automated MR analysis weekly or monthly. In this study, seven different fetal brain tissues were segmented fully-automated with 3D U-Net. The FeTA2021 dataset was divided into the early, middle, and late gestational age groups, different 3D U-Net models were trained for each age group, and segmentation performance was analyzed. Average Dice scores of 0.83, 0.91, and 0.92 were achieved for the gestational-age groups, respectively.

Keywords:

fetal brain tissue, MR imaging, segmentation, U-Net,

PDF

___

S. L. Connors, P. Levitt, S. G. Matthews et al., Fetal mechanisms in neurodevelopmental disorders. Pediatric Neurology, 38, 3, 163–176, 2008. https://doi.org/10.1016/j.pediatrneurol.2007.10.009
A. Barry, B.M. Patten, B.H. Stewart., Possible factors in the development of the Arnold-Chiari malformation. Journal of Neurosurgery, 14, 3, 285–301,1957. https://doi.org/10.3171/jns.1957.14.3.0285
N. E. Foster, K. A. Doyle-Thomas, A. Tryfon et al., Structural gray matter differences during childhood development in autism spectrum disorder: A multimetric approach. Pediatric Neurology, 53, 4, 350–359, 2015. https://doi.org/10.1016/j.pediatrneurol.2015.06.013
Ghai, S., Fong, K.W., Toi, A., Chitayat, D., Pantazi, S., Blaser, S., Prenatal US and MR imaging findings of lissencephaly: review of fetal cerebral sulcal development. Radiographics, 26, 2, 389–405, 2006. https://doi.org/10.1148/rg.262055059
K. Payette, P. de Dumast, H. Kebiri et al., An automatic multi-tissue human fetal brain segmentation benchmark using the fetal tissue annotation dataset. Scientific Data, 8, 1, 1–14, 2021. https://doi.org/10.6084/m9.figshare.14039327
L. Fidon, M. Aertsen, S. Shit et al., Partial supervision for the feta challenge 2021. arXiv preprint https://doi.org/10.48550/arXiv.2111.024082111.02408,2021
P. de Dumast, H. Kebiri, K. Payette, A. Jakab, H. Lajous and M. B. Cuadra, Synthetic magnetic resonance images for domain adaptation: Application to fetal brain tissue segmentation. 2022 IEEE 19th International Symposium on Biomedical Imaging (ISBI) IEEE, pp. 1–5, Kolkata, India, 2022. https://doi.org/10.1109/ISBI52829.2022.9761451
H. Lajous, C. W. Roy, T. Hilbert et al., A fetal brain magnetic resonance acquisition numerical phantom (fabian). Scientific Reports, 12, 1, 1–21, 2022. https://doi.org/10.1038/s41598-022-10335-4
D. Karimi, C. K. Rollins, C. Velasco-Annis, A. Ouaalam and A. Gholipour, Learning to segment fetal brain tissue from noisy annotations. Medical Image Analysis, 85, 102731, 2023. https://doi.org/10.1016/j.media.2022.102731
S. S. M. Salehi, S. R. Hashemi, C. Velasco-Annis et al., Real-time automatic fetal brain extraction in fetal mri by deep learning. 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018), pp. 720-724, Washington, DC, USA, 2018. https://doi.org/10.1109/ISBI.2018.8363675
S. Tourbier, X. Bresson, P. Hagmann, J. P. Thiran, R. Meuli, M.B. Cuadra, An efficient total variation algorithm for super-resolution in fetal brain MRI with adaptive regularization. Neuroimage, 118, 584-97, 2015. https://doi.org/10.1016/j.neuroimage.2015.06.018
M. Kuklisova-Murgasova, G. Quaghebeur, M.A. Rutherford, J.V. Hajnal, J.A. Schnabel, Reconstruction of fetal brain MRI with intensity matching and complete outlier removal. Medical Image Analysis, 16, 8, 1550-64, 2012. https://doi.org/10.1016/j.media.2012.07.004
L. Fidon, M. Aertsen, D. Emam et al., Label-set loss functions for partial supervision: Application to fetal brain 3d mri parcellation. International Conference on Medical Image Computing and Computer-Assisted Intervention Springer, pp. 647–657, Virtual Event, 2021. https://doi.org/10.1007/978-3-030-87196-3_60
A. Makropoulos, E. C. Robinson, A. Schuh et al., The developing human connectome project: A minimal processing pipeline for neonatal cortical surface reconstruction. Neuroimage, 173, 88–112, 2018. https://doi.org/10.1016/j.neuroimage.2018.01.054
Ö. Çiçek, A. Abdulkadir, , S.S. Lienkamp, T. Brox, O. Ronneberger, 3D U-Net: Learning dense volumetric segmentation from sparse annotation. Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2016, pp. 424-432, Athens, Greece, 2016. https://doi.org/10.1007/978-3-319-46723-8_49
Y. Wu, K. He, Group Normalization. Computer Vision - ECCV 2018, pp. 3-19, Munich, Germany, 2018. https://doi.org/10.1007/978-3-030-01261-8_1
L.N. Smith, Cyclical learning rates for training neural networks. 2017 IEEE Winter Conference on Applications of Computer Vision (WACV), pp. 464-472, Santa Rosa, CA, USA, 2017. https://doi.org/10.1109/WACV.2017.58