3D Anatomy of the Atrioventricular Conduction Axis Reconstructed Relative to Gross Anatomical Landmarks Using Hierarchical Phase-contrast Tomography

Hierarchical phase-contrast tomography (HiP-CT) provides 3D imaging at micron-level spatial resolution.¹ Using this technique to determine the arrangement of the atrioventricular conduction axis, we showed recently that the findings validated the accounts that had been provided at the turn of the 20th century.^2–4 Despite the fact that the initial accounts have been shown to be limited by the 2D nature of histology, our study showed that it was possible to predict the likely location of the conduction axis in vivo by CT.⁵ However, the basis of our predictions had been the illustration of the arrangement of the axis as provided by Tawara in his initial study.⁶

We have now resampled the HiP-CT dataset from its original 79.56 µm spatial resolution to 147.5 µm spatial resolution, allowing segmentation and volume rendering with the clinical software used to reconstruct our in vivo datasets. In this way, we have been able manually to delineate the 3D anatomy of the conduction axis, and to demonstrate its location within the autopsied heart. The heart dataset was from body donor S-20-29, an 80-year-old man with history of arterial hypertension, type 2 diabetes and kidney failure. The heart had been scanned at 19.89 µm at the European Synchrotron Radiation Facility, with the dataset now available in open-access format.⁷ Volume-rendered 3D images of the heart were created using Ziostation2 (Ziosoft USA).

We have previously described three histological criteria used to delineate specialised conducting tracts within the myocardium, methods used to manually segment the atrioventricular conduction system in the present investigation.^2,3,5 Although it was possible to identify the more distal fascicles, we were limited in identifying their smaller branches. Assessment of the rendered conduction axis relative to adjacent cardiac structures (Figure 1) now confirms the accuracy of the arrangements predicted on the basis of the initial histological findings of Tawara.⁶

The findings from our initial study have already been used to enhance attempts to locate the conduction axis by clinical imaging.⁸ The similarity between our current direct segmentation of the axis and the depiction offered by Tawara⁶ is remarkable. Nonetheless, we are aware that the axis is not always disposed in the fashion currently demonstrated. Interrogation of additional datasets using this technique will surely enhance the ability of clinicians to use in vivo CT datasets so as better to help them as they aim to either avoid or target the conduction system.⁹