Paper number: 136 A deep learning approach to segmentation of the developing cortex in fetal brain MRI with minimal manual labeling AE Fetit , A Alansary, L Cordero-Grande, J Cupitt, AB Davidson, AD Edwards, JV Hajnal, E Hughes, K Kamnitsas, V Kyriakopoulou, A Makropoulos, PA Patkee, AN Price, MA Rutherford, D Rueckert
Context – Developmental Brain Mapping The Developing Human Connectome Project (DHCP) aims to make major scien;fic progress by crea;ng the first 4D connectome map of early life.
Segmenta5on – Ul5mate Goal Develop a 3D structural segmenta;on pipeline for fetal brain MRI to support connectomics research.
Segmenta5on - Challenges • Rapid changes in morphology over narrow ;me-scales. • Changes in white/grey-maKer intensi;es also take place.
Deep Learning? Successful in other medical imaging applica;ons • However, main difficulty is in the need for large annotated ground-truth. • Whilst large public datasets exist, they tend to mainly include adult brain scans e.g. UK Biobank.
Minimal Labeling Workflow Apply Draw-EM with fetal atlas to generate preliminary 3D labels Fine-tune the 3D Train a 3D cortex Manual QC cortex segmentation segmentation CNN CNN Train a multiclass 3D CNN using scans that passed QC step Apply the multiclass 3D CNN Give ~300 2D slices to Refine the cortex expert annotator labels
CNN Architecture • 3D DeepMedic 3D mo modeling deling u usin sing g DeepMedic t al. 2016 Ka Kamnitsas e et al. 2016 • Th Three parallel pathways: • no normal rmal reso solu;o lu;on n • do downsample wnsampled b by 3 y 3 • do downsample wnsampled b by 5 y 5 • 8 la 8 layer ers per pa s per path thway y • Tr Training batch size was set to 5 • Learning r e-defined schedule. Learning rate f e follo llowed a pr ed a pre-defined schedule.
Preliminary mul5class labels
Example cortex refinement Gestational age: 27.5 weeks
Example cortex segmenta5on Gestational age: 28 weeks
Paper number: 136 Thank you! Questions?
Recommend
More recommend
