Fraunhofer Institute for Digital Medicine MEVIS

Bild IntelliCath mit Torben Pätz und Jan Strehlow

Software Assistance for Endovascular Interventions

Fraunhofer MEVIS develops software solutions for endovascular interventions, especially endovascular navigation.

Clinical Challenges

Reaching the target region during endovascular procedures like angioplasty or embolization is challenging. Within a constrained timeframe, the clinician has to navigate through a tortuous vasculature with numerous bifurcations, largely based on information gathered from 2D angiographic projections of the 3D vasculature. Consequently, clinical studies report long endovascular navigation times, significant ionizing radiation doses, and a significant amount of unhealthy contrast agents used during the navigation. We develop tools for path planning, vessel assessment, and intraoperative support, such as 3D navigation that overcome these challenges.

Highlights

Disruptive 3D device tracking technologies
Customizable real-time capable software platform
Model-based 3D navigation and therapy visualization
Prepared for deliveries according to ISO 62304

Technologies

Software Assistance

Fraunhofer MEVIS develops software solutions for image-guided interventions. By analysis of planning images, patient-specific models, intra-interventional guidance, and post-interventional quantitative assessment, safer and more efficacious image-guided interventions will be possible.

Tracking Systems for Endovascular Devices

Fiber optic device tracking and electromagnetic tracking enable endovascular interventions without ionizing radiation. Both approaches allow a 3D localization of devices and may compensate patient motion.

Vessel Segmentation

DeepLearning-based segmentation of the vascular system from CTA or MRA images provides a road map that is used before and during the intervention for path planning, vessel assessment, and intra-interventional information fusion.

Image Segmentation

Path Planning

By planning the pathway from the insertion point of the endovascular device to the target, we support device selection, path examination, and detection of deviations from the planned path during navigation.

Image-Based Endovascular Device Tracking

Detection and tracking of endovascular devices in angiographic images via DeepLearning provides real-time position feedback. The fusion of this tracking information with diagnostic images and planning information enables navigation support.

Augmented Reality

Combining advanced data processing and visualization with state-of-the-art augmented reality devices for visual feedback and interaction enables smart interventions that provide the clinician with the right information at the right time.

Nav EVAR

2D-3D-Registration

Fusing the angiographic real-time images with diagnostic 3D images enables a transfer of planning information into the interventional scenario or a backprojection of the tracked endovascular device into the planning data.

Real-Time Processing

Various processing tasks, like tracking, registration and visualization, have to be executed on the real-time intra-interventional images. Our dedicated real-time processing framework supports parallel asynchronous processing tasks and combination of results in a modular fashion.

Vessel Assessment

We quantify disease parameters such as stenosis-grade and aneurysms-size in the target vessels as well as along the endovascular pathway to enable a thorough intervention preparation.

Auditory Display

Moving navigation feedback from the visual to the audio channel relieves the clinician from a visual information overflow during endovascular interventions. Based on the planned path and device tracking, auditory display provides information about distances to bifurcations, required device rotations, and whether the device is on track.