Towards Real-Time Autonomous Navigation: Transformer-Based Catheter Tip Tracking in Fluoroscopy

TL;DR

This paper presents a real-time fluoroscopy-based catheter tip tracking system using deep learning, enabling autonomous robotic navigation in stroke treatment with improved accuracy and robustness.

Contribution

Developed a multi-threaded deep learning pipeline with segmentation models that outperform existing benchmarks for real-time catheter tip tracking in fluoroscopy.

Findings

SegFormer achieved a mean absolute error of 4.44 mm, outperforming other models.

System exceeded state-of-the-art CathAction results with up to +5% Dice score improvements.

Framework maintains stable performance under challenging fluoroscopic imaging conditions.

Abstract

Purpose: Mechanical thrombectomy (MT) improves stroke outcomes, but is limited by a lack of local treatment access. Widespread distribution of reinforcement learning (RL)-based robotic systems can be used to alleviate this challenge through autonomous navigation, but current RL methods require live device tip coordinate tracking to function. This paper aims to develop and evaluate a real-time catheter tip tracking pipeline under fluoroscopy, addressing challenges such as low contrast, noise, and device occlusion. Methods: A multi-threaded pipeline was designed, incorporating frame reading, preprocessing, inference, and post-processing. Deep learning segmentation models, including U-Net, U-Net+Transformer, and SegFormer, were trained and benchmarked using two-class and three-class formulations. Post-processing involved two-step component filtering, one-pixel medial skeletonization, and greedy arc-length path following with contour fall-back. Results: On manually-labeled moderate complexity fluoroscopic video data, the two-class SegFormer achieved a mean absolute error of 4.44 mm, outperforming U-Net (4.60 mm), U-Net+Transformer (6.20 mm) and all three-class models (5.19-7.74 mm). On segmentation benchmarks, the system exceeded state-of-the-art CathAction results with improvements of up to +5% in Dice scores for three-segmentation. Conclusion: The results demonstrate that the proposed multi-threaded tracking framework maintains stable performance under challenging imaging conditions, outperforming prior benchmarks, while providing a reliable and efficient foundation for RL-based autonomous MT navigation.