The investigator's project proposes the development of a 3D hybrid guidance system which has the aim of avoidance of scar and septal perforation through targeted lead deployment via a personalised septal real time image overlay onto x-ray fluoroscopy imaging during left bundle branch pacing. The investigators hypothesise that the use of cardiac anatomy and myocardial scar distribution derived from cardiac magnetic resonance imaging (MRI) as well as 3D position of the pacing lead, may improve LBBAP lead deployment success and improve clinical outcomes by guiding the physician towards optimal lead positioning.
The study will use anonymised imaging data from the TACTIC-CRT (IRAS 250715) , Cardiac CT to guide Cardiac Resynchronisation Therapy (CRT) implantation (IRAS 150161) study (IRAS 150161) and anonymous cardiac MRI, CT and fluoroscopy imaging data from patients who underwent LBBA pacing to develop a 3D hybrid guidance system using machine learning methods. The developed 3D hybrid guidance system will be tested on anonymised retrospectively collected x-ray fluoroscopy images from patients who underwent LBBAP and had a cardiac MRI. The 3D hybrid guidance system will collect data from the retrospective anonymised X-ray fluoroscopic images, detect the pacing lead and reconstruct the 3D position of the pacing lead and test the accuracy of the 3D hybrid guidance system.
Study Type
OBSERVATIONAL
Enrollment
300
This is a retrospective observational study using cardiac MRI, cardiac CT and x-ray fluoroscopy images to develop a 3D hybrid guidance system with the use of deep learning methods.
Imperial College Nhs Healthcare Trust
London, United Kingdom
Detection of the position of the left bundle branch area on x-ray fluoroscopy
The accuracy of left bundle branch area detection will be measured as intersection-over-union (IoU) with the use of a convolutional neural network
Time frame: 12 months
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.