4 | DISCUSSION
TEE is the primary image guidance modality used for trans-catheter
structural heart intervention technologies because of its better
real-time visualization of cardiac anatomical structures. However, its
limited sector width and complex probe orientation prolong the learning
curve to comprehend the spatial configuration of procedural TEE imaging.
Although CT has been widely used as a preoperative planning tool for
trans-catheter structural heart interventions, the created 2D CT images
are usually taken from different visual perspectives of intraoperative
TEE images. This creates issues in positioning, docking and taking other
specific key steps of these image-guided interventions, because the
procedural TEE imaging may fail to reproduce the same prerequisite
patterns as the CT based preoperative planning. Compared to the high
degree of freedom in the CT MPR slices, the motion of TEE probe is
rather limited. Therefore, the methodology which we propose here is
capable of carrying out a preoperative plan by CT MPR to simulate the
intraoperative TEE rather than applying the TEE to acquire images
conforming to what is showed on CT slices. In this article, we
demonstrate a stepwise CT MPR manipulating method to simulate the TEE
probe motion and preview the procedural TEE images on a different
imaging modality. This reduces somewhat the differences in deciding the
size of devices from these two imaging modalities. Besides validating
this method in different TEE views, we also applied it on dynamic CT to
mimic TEE videos. This concept has not previously been introduced in a
systematic fashion.
On the other hand, such a learning curve is also steep for
interventional echocardiographers to precisely offer the
procedure-specific imaging needs in every patient. 13Although basic acquisition protocols of TEE have been introduced by Hahn
and colleagues, 13–16 many patient-specific
anatomical conditions, such as clockwise/counterclockwise rotation of
the heart, trajectory of esophageal tract, geometry of gastric space and
its relative position with cardiac structures all lead to significant
differences in the spatial configuration of procedural TEE imaging. As a
result, the TEE imaging is of large variability and highly technique
dependent. Sometimes, it is even impossible to obtain the ideal
procedure-specific views in patients of challenging anatomy. In the
application of a traditional method, interventional echocardiographers
discern not the feasibility of those procedure-specific views until TEE
probe is inserted. Fortunately, in our simulation, the CT datasets in
nature contain all the anatomical information, and thus broaden the
vision of interventional echocardiographers beyond the scanning sector,
offer the spatial orientation between esophagus and the interested
cardiac structure, and thus enable us to foresee whether the desired
procedure-specific imaging is feasible or not since each step in our
method can be translated back into a patient-specific TEE acquisition
protocol. If the procedure-specific imaging is not feasible, the
heart-team can promptly use our method to seek surrogate TEE views at
other esophageal or transgastric levels. This assists interventional
echocardiographers and operators reach to a consensus in the
preoperative rehearsal and may even shorten the procedure time and
fluoroscopic time.
In this era of echocardiography-fluoroscopic and CT-fluoroscopic fusion
imaging, one might underestimate the importance in comprehending the 3D
orientation of 2D TEE images. Nevertheless, it is important for
transcatheter interventionists to extend their proprioception all the
way to the catheter tip and nearby structures. Proprioception needs
visual assistance to become precise perception; however, unlike the
electroanatomical mapping systems of the ablation catheters, current
devices for structural heart diseases are still unable to work on
precise targeting or tracking in the 3D space but only projecting to the
fluoroscopy where essential spatial resolution is lacking. In fact,
simply putting all modalities onto a cine fluoroscopy produces nothing
more than confusions because there are either just static segmentations
or mere artificial intelligence based modeling rather than imaging.
Although fusion imaging with high-resolution TEE images is feasible
nowadays, anatomical details are still discernible only when the image
intensifier is at the correct projection angles to spread the 2D images.
Our method allows patient-specific fluoroscopic viewing angles to be
generated simultaneously during the MPR manipulation. In combination
with the concept of the optimal projection curve and the
projection-slice relationship, 9,10 CT MPR can enhance
a better understanding of the spatial orientation of
echocardiography-fluoroscopic fusion imaging.