Dual Robot System for Autonomous Needle Insertion into Deep Vessels

Precise needle insertion is a key operation in many medical procedures such as peripheral catheterization, cardiac endovascular treatments, biopsy and treatment of tumours in soft tissues such as breast, prostate and abdomen [1]. Among them, Central Venous Access (CVA) is a routine procedure typically performed by experienced clinicians under ultrasound or x-ray guidance in a surgical room environment. CVA is most commonly conducted in three deep vein locations: the internal jugular vein, the subclavian vein, and the femoral vein. During CVA, clinician should insert the needle while maintaining proper visualization of the target vessel (i.e. ultrasound-guided access). However, performing the de- tection and continuous visualization of the target vessel and simultaneously precisely controlling the needle inser- tion is not trivial. At the same time, complication rates during CVA range up to 15% (mechanical in 5-19% of patients; infectious in 5-26%; thrombotic in 2-26%) [2]. In this paper, we propose a dual robot system for autonomous needle insertion into deep vessels by enabling real-time visualization of the vein and adaptive trajectory planning to provide safe and quick interactions.