Current techniques to image the microstructure of the heart with diffusion tensor MRI (DTI) are highly under-resolved. We present a technique to improve the spatial resolution of cardiac DTI by almost 10-fold and leverage this to measure local gradients in cardiomyocyte alignment or helix angle (HA). We further introduce a phenomapping approach based on voxel-wise hierarchical clustering of these gradients to identify distinct microstructural microenvironments in the heart. Initial development was performed in healthy volunteers (n=8). Thereader, subjects with severe but well-compensated aortic stenosis (AS, n=10) were compared to age-matched controls (CTL, n=10). Radial HA gradient was significantly reduced in AS (8.0±0.8°/mm vs. 10.2±1.8°/mm, p=0.001) but the other HA gradients did not change significantly. Four distinct microstructural clusters could be idenJfied in both the CTL and AS subjects and did not differ significantly in their properties or distribution. Despite marked hypertrophy, our data suggest that the myocardium in well-compensated AS can maintain its microstructural coherence. The described phenomapping approach can be used to characterize microstructural plasticity and perturbation in any organ system and disease.

中文翻译：

---

健康和压力超载的人类心脏的扩散张量表观映射

目前使用扩散张量 MRI (DTI) 对心脏微观结构进行成像的技术分辨率严重不足。我们提出了一种将心脏 DTI 的空间分辨率提高近 10 倍的技术，并利用该技术来测量心肌细胞排列或螺旋角 (HA) 的局部梯度。我们进一步介绍了一种基于这些梯度的体素层次聚类的表型图方法，以识别心脏中不同的微结构微环境。最初的开发是在健康志愿者（n=8）中进行的。读者将患有严重但代偿良好的主动脉瓣狭窄的受试者（AS，n = 10）与年龄匹配的对照者（CTL，n = 10）进行比较。 AS 中径向 HA 梯度显着降低（8.0±0.8°/mm 对比 10.2±1.8°/mm，p=0.001），但其他 HA 梯度没有显着变化。在 CTL 和 AS 受试者中都可以识别出四个不同的微观结构簇，并且它们的特性或分布没有显着差异。尽管明显肥厚，我们的数据表明代偿良好的 AS 中的心肌可以保持其微观结构的一致性。所描述的表型图谱方法可用于表征任何器官系统和疾病的微观结构可塑性和扰动。