Hendrickson Timothy J, Reiners Paul, Moore Lucille A, Lundquist Jacob T, Fayzullobekova Begim, Perrone Anders J, Lee Erik G, Moser Julia, Day Trevor K M, Alexopoulos Dimitrios, Styner Martin, Kardan Omid, Chamberlain Taylor A, Mummaneni Anurima, Caldas Henrique A, Bower Brad, Stoyell Sally, Martin Tabitha, Sung Sooyeon, Fair Ermias A, Carter Kenevan, Uriarte-Lopez Jonathan, Rueter Amanda R, Yacoub Essa, Rosenberg Monica D, Smyser Christopher D, Elison Jed T, Graham Alice, Fair Damien A, Feczko Eric
bioRxiv. 2025 Jan 11:2023.03.22.533696. doi: 10.1101/2023.03.22.533696.
Brain segmentation of infant magnetic resonance (MR) images is vitally important for studying typical and atypical brain development. The infant brain undergoes many changes throughout the first years of postnatal life, making tissue segmentation difficult for most existing algorithms. Here we introduce a deep neural network BIBSNet ( aby and nfant rain egmentation Neural work), an open-source, community-driven model for robust and generalizable brain segmentation leveraging data augmentation and a large sample size of manually annotated images.
Included in model training and testing were MR brain images from 90 participants with an age range of 0-8 months (median age 4.6 months). Using the BOBs repository of manually annotated real images along with synthetic segmentation images produced using SynthSeg, the model was trained using a 10-fold procedure. Model performance of segmentations was assessed by comparing BIBSNet, joint label fusion (JLF) inferred segmentation to ground truth segmentations using Dice Similarity Coefficient (DSC). Additionally, MR data along with the FreeSurfer compatible segmentations were processed with the DCAN labs infant-ABCD-BIDS processing pipeline from ground truth, JLF, and BIBSNet to further assess model performance on derivative data, including cortical thickness, resting state connectivity and brain region volumes.
BIBSNet segmentations outperforms JLF across all regions based on DSC comparisons. Additionally, with processed derived metrics, BIBSNet segmentations outperforms JLF segmentations across nearly all metrics.
BIBSNet segmentation shows marked improvement over JLF across all age groups analyzed. The BIBSNet model is 600x faster compared to JLF, produces FreeSurfer-compatible segmentation labels, and can be easily included in other processing pipelines. BIBSNet provides a viable alternative for segmenting the brain in the earliest stages of development.
婴儿磁共振（MR）图像的脑部分割对于研究典型和非典型脑发育至关重要。婴儿大脑在出生后的头几年会经历许多变化，这使得大多数现有算法进行组织分割变得困难。在此，我们引入了一种深度神经网络BIBSNet（婴儿脑部分割神经网络），这是一个开源的、由社区驱动的模型，用于通过数据增强和大量手动标注图像的样本量进行稳健且可推广的脑部分割。
模型训练和测试中纳入了90名年龄在0 - 8个月（中位年龄4.6个月）参与者的MR脑图像。利用手动标注的真实图像的BOBs存储库以及使用SynthSeg生成的合成分割图像，该模型采用10折交叉验证法进行训练。通过使用骰子相似系数（DSC）将BIBSNet、联合标签融合（JLF）推断的分割与地面真值分割进行比较，来评估分割的模型性能。此外，将MR数据以及与FreeSurfer兼容的分割结果通过DCAN实验室的婴儿-ABCD-BIDS处理管道从地面真值、JLF和BIBSNet进行处理，以进一步评估模型在派生数据上的性能，包括皮质厚度、静息态连接性和脑区体积。
基于DSC比较，BIBSNet分割在所有区域均优于JLF。此外，在处理后的派生指标方面，BIBSNet分割在几乎所有指标上均优于JLF分割。
在所有分析的年龄组中，BIBSNet分割均显示出比JLF有显著改进。BIBSNet模型比JLF快600倍，生成与FreeSurfer兼容的分割标签，并且可以轻松纳入其他处理管道。BIBSNet为在发育的最早阶段进行脑部分割提供了一种可行的替代方案。
