Color Coded Angiography: Real Time Quantitative Imaging

Presented at the 15th Biennial International Andreas Gruentzig Society Meeting, February 3-7, 2019

Program Agenda               Faculty Disclosures              Vendor Acknowledgment

3.3  /  IAGS 2019
Session 3: Endovascular Session 1
Color Coded Angiography: Real Time Quantitative Imaging
Eric J. Dippel, MD, FACC


Statement of problem or issue

The goals for treating critical limb ischemia (CLI) are to maximize arterial inflow to the foot, heal wounds, minimize tissue loss, and prevent major amputations. Endovascular techniques and tools have evolved to the point that percutaneous revascularization below the ankle, including pedal arch reconstruction, has become possible, particularly in centers of excellence for CLI therapy. Typically, in the angio suite, we rely on digital subtraction angiography (DSA) to make decisions on procedural success, completeness of revascularization, and predict wound healing.  


Gaps in knowledge

DSA has not evolved much since its advent in the 1980s, and it has a number of limitations, including the fact that interpretation is highly subjective and only qualitative in determining arterial flow. The subjective nature of DSA data precludes analyses that potentially might aid with real-time decision making. Additionally, DSA images are acquired in black and white, but we live in a color world. In today’s practice, an endovascular revascularization is performed on a patient with CLI and they are then sent back to the wound clinic with the hope that an adequate revascularization was performed, but without quantitative data that it was. Historically, there is no objective, predictive in-vivo imaging performed in the angio suite to assess the completeness of the revascularization. Traditional non-invasive surrogate assessment of limb perfusion, such as ankle-brachial index (ABI), Sensilace, or Tcom measurements, are done at some later date after the intervention and therefore not available in real-time. 


Possible solutions and future directions

Parametric angiographic imaging can provide near real-time, objective, quantitative, color-coded assessment of blood flow to the affected limb in-vivo in the angio suite.  The DSA images are post-processed via a high bandwidth server and displayed within seconds of acquisition. The parametric imaging software analyzes every pixel of every frame of an acquisition run and produces a time-density contrast curve for each pixel.  This time density curve can then be used to define discrete data points, such as: (1) time to arrival, (2) time to peak, (3) area under the curve, (4) peak height, and (5) mean transit time. These data points can then be analyzed objectively and quantitatively, potentially using artificial intelligence, to make decisions on the completeness of revascularization in the angio suite prior to ending the case. They can also be mapped to create a color-coded angiogram (Figure 1).

In order to compare pre- and post-intervention parametric images, there are several technical considerations that are important. There must be no motion artifact during image acquisition. The image detector height, the table height, the field of view, the camera magnification, and the camera angulation must all be the same for the before and after images. There should be no compensation filter in the field of view. The pre and post angiograms should be obtained from the same injection site, and the contrast flow rate and volume should be the same. If a vasodilator was used, the same dose should be administered with the pre and post images. Finally, the same region of interest (ROI) should be used for analysis. 

Parametric imaging is in its infancy regarding potential applications for limb salvage.  This has tremendous potential in the treatment of CLI to predict the adequacy of restoration of arterial inflow and corelate to the success or futility of wound healing.  There are a number of future directions that need to be addressed. The standardization of image acquisition is paramount to be able to conduct any meaningful large-scale multicenter studies. The size and location of the ideal ROI must be determined. Normal parameter ranges of the contrast time density curve need to be defined. There should be comparative studies with various vasodilators and doses to determine what is optimal, this also raises the question of whether a vasodilator is even necessary before imaging. An automated algorithm for determining the data points from the contrast time intensity curve would be very useful. Ultimately, the goal is to use this exciting imaging modality seamlessly during an intervention to predict wound healing, restenosis, late target lesion/vessel failure, long-term success/outcome, or even comparatively between different devices.

Figure 3.3