Improving neck ultrasound image retrieval using intra-sweep representation learning.

Purpose: Intraoperative ultrasound (US) can enhance real-time visualization in transoral robotic surgery (TORS) and improve the safety of the surgery. To develop a US guidance system for TORS, US probe localization and US-preoperative image registration are essential. Image retrieval has the potential to solve these two problems in the same framework, and learning a discriminative US representation is key to successful image retrieval.

Methods: We propose a self-supervised contrastive learning approach to match intraoperative US views to a preoperative image database. We introduce a novel contrastive learning strategy that leverages intra-sweep similarity and US probe location to improve feature encoding. Additionally, our model incorporates a flexible threshold to reject unsatisfactory matches.

Results: Our method achieves 92.30% retrieval accuracy on simulated data and outperforms state-of-the-art temporal-based contrastive learning approaches. We also test our approach on real patient data with preoperative US-CT registration to show the feasibility of the proposed US probe localization system, despite tissue deformation due to tongue retraction.

Conclusion: Our contrastive learning method, which utilizes intra-sweep similarity and US probe location, enhances US image representation learning. We also demonstrate the feasibility of using our image retrieval method to provide neck US localization on real patients US after tongue retraction. Total number of words: 2414 words.