Deep learning assisted intraoperative instrument cleaning station for robotic scrub nurse systems

Abstract

Abstract Due to the ongoing shortage of qualified surgical assistants and the drive for automation, the deployment of robotic scrub nurses (RSN) is being investigated. As such robotic systems are expected to fulfill all indirect and direct forms of surgical assistance currently provided by human operating room (OR) assistants, they must also be capable of performing intraoperative cleaning of laparoscopic instruments, which are prone to contamination when using electrosurgical techniques during minimally invasive procedures. We present a cleaning station for robotic scrub nurse systems which provides intraoperative cleaning of laparoscopic instruments during minimally invasive procedures. The system uses deep learning to decide autonomously on the need of intraoperative cleaning to preserve instrument functions. We performed configuration and durability tests to determine an optimal set of system parameters and to verify the system performance in an application context. The results of the configuration tests indicate that the use of hard brushes in combination with a sodium chloride cleaning solution and a sequence of 3 s cleaning intervals provides the best cleaning performance with a minimal total cleaning time. The results of the durability tests show that the cleaning function is in principle guaranteed for the duration of a surgical intervention. Our evaluation tests have shown that our deep learning assisted cleaning station for robotic scrub nurse systems is capable of performing autonomous intraoperative cleaning of laparoscopic instruments, providing a further step towards the integration of robotic scrub nurse systems into the OR.