What is Endoleak Risk Index (ERI)?
What is Endoleak Risk Index (ERI)?
Endoleak Risk Index, is an AI-powered index based on the analysis of the intervention simulation, with a specific focus on the proximal apposition between the endograft and the patient-specific aortic digital twin. It assesses preoperatively the risk of Type IA endoleak associated with EVAR.
How is Endoleak Risk Index (ERI) calculated?
Each simulation provides detailed information about the interactions between the stent-graft and the patient-specific digital twin. From a single EVAR simulation, thousands of data points can be extracted.
In fact, the following methodology is used :
Β·Β Β Analysis of the proximal neck: Transversal slices are automatically performed every 0.5 mm on the 2 centimeters of the aortic neck with best apposition, resulting in 40slices.Β
Β·Β Β Discretisation of Graft and Aortic wall circumferences: For each slice, both the aortic wall and the endograft are discretised into 200 points each.Β
Β·Β Β Radius measurements: For each point of the aortic wall (excluding thrombus and calcifications), the distance to the center of the aortic lumen is calculated. Similarly, for each point of the endograft, the distance to the center of the aortic lumen is also measured.
If we look at one particular transversal slice, these radius measurements of both the aortic wall and the endograft provide local information on apposition and oversizing of the graft, as well as detailed morphology of the aortic neck, such as conicity, or circularity.Β
When combined, this process produces 400 radius measurements per slice (200 for the aorta and 200 for the endograft). Across 40 slices, this results in 16,000 measurements per patient.
Are thrombus and calcification taken into account?
It is important to distinguish between two different stages of analysis:Β β
During Digital twin simulation: Only the aortic lumen is modeled, Thrombus and calcifications are not currently included in the simulated geometry. This represents a known limitation of the present technology, and research is ongoing to incorporate these structures into future product versions.
During ERI risk computation: While thrombus is not directly modelled in the simulation, its clinical relevance is not ignored. The amount of thrombus in the aortic neck is included as a variable in the machine learning algorithm used to compute the ERI. This approach ensures that thrombus burdenβ is taken into account when generating the final risk assessment.
How does the Endoleak Risk Index (ERI) take into account the way the operator will deploy the stent-graft?
Proximal landing zone of the stent graft is actually an input defined by the clinical specialist or the physician as a distance to a reference artery, usually the lower edge of the lowest renal artery. If the operator is not sure about the landing zone reachable, multiple proximal landing zones scenarios can be simulated.