Active Search of Subsurface Lymph Nodes Using Robot-Assisted Electrical Impedance Scanning

Abstract

Lymphadenectomy is frequently performed for cancer treatment. Since lymph nodes are surrounded by fatty tissues, they are often difficult to detect. In robotic-assisted minimally invasive surgery (RMIS), the difficulty increases because haptic feedback is unavailable. This article presents a novel sensing system to assist surgeons with subsurface lymph node detection. The proposed system uses already existing instruments for measuring tissues&#x2019; electrical properties. A machine-learning-based classifier is developed to estimate the likelihood of a lymph node being present at the measuring site. In addition, an optimized area search algorithm is integrated to make the sensing procedure autonomous and efficient. The proposed system is built and evaluated through experiments using water tank setups, finite element simulation, and <i>ex vivo</i> tissue phantoms. The results demonstrate the efficacy of the proposed method including high detection precision, recall, and Matthews correlation coefficient (MCC). Besides, the proposed method can greatly reduce the number of sampling points compared with a grid-based search method, leading to a quarter faster for the acquisition time. Given the promising performance and easy implementation, the proposed system can potentially improve the quality of related surgical procedures significantly in the future.