Introduction

Real-time MR-thermometry shows promise for the guidance of cardiac ablation procedures^1,2. Active ablation catheters, containing tracking coils, are currently used for these procedures³. During ablation in the heart, the catheter location is subject to cardiac and respiratory motion and can potentially move/drift during the procedure, resulting in inaccurate lesion formations. Very few studies have investigated the impact of catheter drift⁴, which is difficult to identify during cardiac MR-thermometry. In this study, we sought to develop a catheter drift detection technique within a cardiac MR-thermometry sequence for real-time simultaneous assessment of catheter drift and lesion formation.

Methods

Proposed approach
A continuous acquisition of active tracking (AT) modules was added to a cardiac EPI MR-thermometry sequence, as previously demonstrated⁵ and shown in Figure 1. This allows AT modules to be acquired repeatedly at a sampling rate of 24ms (42Hz) to measure the catheter position. The temporal evolution of the AT signal was used for two purposes. First, it was employed to detect a robust cardiac trigger for the MR-thermometry acquisition using band-pass filtering to isolate the cardiac motion component of the temporal AT signal (minimum frequency of 0.67Hz and maximum frequency of 2.5Hz), as previously demonstrated⁵. Second, low-pass filtering of the temporal AT signal was applied to remove both cardiac and respiratory motion components and to isolate the potential catheter drift. This was done using a lowpass finite impulse response filter, designed in MATLAB (v2021a, The MathWorks, Nantucket, MI) with a cutoff frequency of 0.05Hz and filter order of 400.

Experimental evaluation
All experiments were performed on a 1.5T scanner (MAGNETOM Aera, Siemens Healthcare, Erlangen Germany). The proposed approach was evaluated in a phantom which was built for these experiments and consisted of a plastic tube with a closed end which was fixed to the bottom of a container filled with water. The catheter (Vision-MR Ablation Catheter, Imricor Medical Systems, Burnsville, MN) was inserted into the tube which restricted its movement to the foot-head direction. The closed end of the tube also ensured that the catheter was returned to the same position after each movement. This phantom set up allowed the catheter to be moved independently from the box. The frequencies at which the catheter and box were moved by were approximated by displacing them continuously by hand in the scanner by a set distance in a known time period. In this study, the catheter was moved at 0.25Hz with an amplitude of 0.5cm (setup #1) and 1.5cm (setup #2) to represent the breathing motion. The box was kept still for the first 20 seconds of each experiment. After that, a permanent foot-head drift combined with a periodic continuous foot-head displacement of 1cm at 1Hz to simulate the cardiac motion was manually applied to the box. Control experiments where the box was stationary for the entire duration of the scan were also carried out.

Results

Figure 2 shows an example of a magnitude image and associated temperature map obtained during the experiments. Figure 3 shows the AT signal acquired during the experiments when the catheter was moved with an amplitude of 1.5cm and 0.5cm (on left and right of Figure3b respectively), together with the corresponding low-filtered signal representing the drift contribution. In both experiments, a drift can be easily identified on the low-pass filtered signals. The cardiac component of the AT signal used for cardiac triggering was also depicted for both experiments (Figure 3c) where noise like signal is observed during the first 20s and is then followed by periodic displacement when the box starts moving to replicate cardiac motion.

Discussion

Catheter drift detection within an MR-thermometry sequence is possible and was successfully demonstrated in phantom experiments. Its in-vivo feasibility and accuracy during an ablation remains to be investigated. In this simulated environment, the drift was only detected in one orientation as the cardiac and respiratory motion were both applied in the foot-head direction. Filtering on a composite metric representing the 3D displacement⁵ may therefore be needed in vivo where catheter drift may occur in any dimension. Additionally, there is a delay that is induced by the filter which may cause a latency in detecting the catheter drift in real time and could be compensated with more advanced filtering strategies.

Conclusion

The proposed MR-thermometry approach shows promise for simultaneous lesion assessment and detection of the catheter drift in the presence of respiratory and cardiac motion. Further studies evaluating the in vivo capabilities of this technique are now warranted.

Acknowledgements

This work was supported by the Innovate UK grant (68539), the Engineering and Physical Sciences Research Council (EPSRC) grant (EP/R010935/1), the British Heart foundation (BHF) grants (PG/19/11/34243 and PG/21/10539), the Wellcome EPSRC Centre for Medical Engineering at Kings College London (WT 203148/Z/16/Z), the National Institute for Health Research (NIHR) Biomedical Research Centre based at Guy’s and St Thomas’ NHS Foundation Trust and King’s College London. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.