6DoF catheter detection, application to intracardiac echocardiography

Area of Science:

• Medical Imaging
• Interventional Cardiology
• Robotics

Background:

• Hybrid imaging systems combining fluoroscopy and echocardiography are vital for minimally invasive cardiac interventions.
• Intracardiac echocardiography (ICE) offers a potential alternative to transesophageal echocardiography (TEE), reducing the need for anesthesia.

Purpose of the Study:

• To introduce and evaluate a novel 6 degrees of freedom (DoF) pose estimation method for catheters using single X-ray fluoroscopy.
• To apply this method to a prototype ICE catheter for improved intra-operative guidance.

Main Methods:

• A machine learning-based catheter detection approach integrated into a Bayesian hypothesis fusion framework.
• Refinement of radiopaque ball marker locations using template matching and iterative optimization for 3D pose estimation.
• Registration of the ICE volume to the C-arm coordinate system.

Main Results:

• The novel pose estimation method was tested on synthetic and porcine in-vivo data.
• The system achieved a preliminary target registration error (TRE) of 8.06 ± 7.2 mm over 703 cases.
• Successful registration of the ICE volume within the C-arm coordinate system.

Conclusions:

• The developed method enables accurate 6 DoF pose estimation of catheters from single X-ray fluoroscopy projections.
• This hybrid imaging approach shows promise for enhancing precision in structural heart disease interventions and electrophysiology procedures.