In the Malvaceae family, dynamic solar tracking by leaves is actuated by the deformation of the pulvinus, a thickened region at the leaf blade–petiole junction. While the internal structure is believed to play a crucial role in this process, experimental verification has been challenging due to technical limitations. To address this gap, we developed a semi-automated workflow, which integrates data analysis and image processing to simultaneously analyze the shape and internal structure of a Malvaceae pulvinus using X-ray microtomography. Firstly, we found that kenaf (Hibiscus cannabinus L.), a Malvaceae species with curved pulvini, exhibited solar-tracking leaf movement and selected it as a model system. We employed diffusible iodine-based contrast-enhanced computed tomography to visualize the internal structure of the kenaf pulvinus. Analysis of the pulvini’s shape revealed variations in pulvinus morphology, yet plausible prediction of the centerline was accomplished using polar polynomial regression. Upon slicing the pulvini perpendicular to the centerline, we observed distinct gray value gradients along the proximo-distal and adaxial-abaxial axes, challenging threshold-based tissue segmentation. This workflow successfully generated three modified 3D images and derived quantitative parameters. Using these quantitative parameters, we conducted network analysis and found the linkage between the size-normalized cortex cross-sectional area and curvature. Polynomial least absolute shrinkage and selection operator (LASSO) regression revealed the relationship between the size-normalized cortex cross-sectional area and curvature commonly in all three tested samples. This workflow enables simultaneous analysis of the shape and internal structure, significantly improving the reproducibility of Malvaceae leaf pulvinus characterization.

在锦葵科中，叶子的动态太阳追踪是由叶枕（叶片与叶柄交界处的增厚区域）的变形驱动的。虽然内部结构被认为在此过程中发挥着至关重要的作用，但由于技术限制，实验验证一直具有挑战性。为了解决这一差距，我们开发了一种半自动化工作流程，它集成了数据分析和图像处理，使用 X 射线显微断层扫描同时分析锦葵科的形状和内部结构。首先，我们发现红麻（Hibiscus cannabinus L.）是一种具有弯曲叶脉的锦葵科物种，表现出跟踪太阳的叶子运动，并选择它作为模型系统。我们采用基于扩散碘的对比增强计算机断层扫描来可视化洋麻枕的内部结构。对枕骨形状的分析揭示了枕骨形态的变化，但使用极多项式回归完成了中心线的合理预测。在垂直于中心线切割枕骨后，我们观察到沿近远轴和近轴-远轴的不同灰度值梯度，这对基于阈值的组织分割提出了挑战。该工作流程成功生成了三个修改后的 3D 图像和导出的定量参数。利用这些定量参数，我们进行了网络分析，发现了尺寸归一化皮层横截面积和曲率之间的联系。多项式最小绝对收缩和选择算子（LASSO）回归揭示了所有三个测试样本中尺寸归一化皮层横截面积和曲率之间的关系。该工作流程可以同时分析形状和内部结构，显着提高锦葵科叶枕特征的再现性。