Kwon Hyeokjin, Son Seungyeon, Morton Sarah U, Wypij David, Cleveland John, Rollins Caitlin K, Huang Hao, Goldmuntz Elizabeth, Panigrahy Ashok, Thomas Nina H, Chung Wendy K, Anagnostou Evdokia, Norris-Brilliant Ami, Gelb Bruce D, McQuillen Patrick, Porter George A, Tristani-Firouzi Martin, Russell Mark W, Roberts Amy E, Newburger Jane W, Grant P Ellen, Lee Jong-Min, Im Kiho
Department of Electronic Engineering, Hanyang University, Seoul, South Korea; Fetal Neonatal Neuroimaging and Developmental Science Center, Boston Children's Hospital and Harvard Medical School, Boston, MA, USA; Division of Newborn Medicine, Boston Children's Hospital, Boston, MA, USA; Department of Pediatrics, Harvard Medical School, Boston, MA, USA.
Department of Artificial Intelligence, Hanyang University, Seoul, South Korea.
Med Image Anal. 2025 May;102:103538. doi: 10.1016/j.media.2025.103538. Epub 2025 Feb 28.
Examining the altered arrangement and patterning of sulcal folds offers insights into the mechanisms of neurodevelopmental differences in psychiatric and neurological disorders. Previous sulcal pattern analysis used spectral graph matching of sulcal pit-based graph structures to assess deviations from normative sulcal patterns. However, challenges exist, including the absence of a standard criterion for defining a typical reference set, time-consuming cost of graph matching, user-defined feature weight sets, and assumptions about uniform node distribution. We developed a deep learning-based sulcal pattern analysis to address these challenges by adapting prototype-based graph neural networks to sulcal pattern graphs. Additionally, we proposed a prototype inverse-projection for better interpretability. Unlike other prototype-based models, our approach inversely projects prototypes onto individual node representations to calculate the inverse-projection weights, enabling efficient visualization of prototypes and focusing the model on selective regions. We evaluated our method through a classification task between healthy controls (n = 174, age = 15.4 ±1.9 [mean ± standard deviation, years]) and patients with congenital heart disease (n = 345, age = 15.8 ±4.7) from four cohort studies and a public dataset. Our approach demonstrated superior classification performance compared to other state-of-the-art models, supported by extensive ablative studies. Furthermore, we visualized and examined the learned prototypes to enhance understanding. We believe our method has the potential to be a sensitive and understandable tool for sulcal pattern analysis.
研究脑沟褶皱排列和形态的改变，有助于深入了解精神疾病和神经疾病中神经发育差异的机制。以往的脑沟模式分析采用基于脑沟凹点的图谱结构的谱图匹配来评估与标准脑沟模式的偏差。然而，仍存在一些挑战，包括缺乏定义典型参考集的标准准则、图谱匹配耗时成本高、用户定义的特征权重集以及关于节点均匀分布的假设。我们开发了一种基于深度学习的脑沟模式分析方法，通过将基于原型的图神经网络应用于脑沟模式图来应对这些挑战。此外，我们还提出了一种原型反投影方法，以提高可解释性。与其他基于原型的模型不同，我们的方法将原型反向投影到单个节点表示上，以计算反投影权重，从而实现原型的高效可视化，并使模型专注于选择性区域。我们通过四项队列研究和一个公共数据集中健康对照组（n = 174，年龄 = 15.4 ±1.9 [平均值±标准差，岁]）和先天性心脏病患者（n = 345，年龄 = 15.8 ±4.7）之间的分类任务对我们的方法进行了评估。在广泛的消融研究支持下，我们的方法与其他先进模型相比表现出了卓越的分类性能。此外，我们对学习到的原型进行了可视化和检查，以增强理解。我们相信我们的方法有潜力成为一种用于脑沟模式分析的灵敏且易于理解的工具。
