Behavioral characteristics of manual palpation to localize hard nodules in soft tissues.

Abstract

Improving the effectiveness of artificial tactile sensors for soft-tissue examination and tumor localization is a pressing need in robot-assisted minimally invasive surgery. Despite the availability of tactile probes, guidelines for optimal palpation behavior that best exploit soft-tissue properties are not available as yet. Simulations on soft-tissue palpation show that particular stress-velocity patterns during tissue probing lead to constructive dynamic interactions between the probe and the tissue, enhancing the detection and localization of hard nodules. To the best of our knowledge, this is the first attempt to methodically evaluate the hypothesis that specific human palpation behaviors (defined by the fingers' velocity, trajectory, and exerted force) directly influence the diagnosis of soft-tissue organs. Here, we use simulation studies involving human participants to establish open hypotheses on the interaction and influence of relevant behavioral palpation variables, such as finger trajectory, its velocity, and force exerted by fingers on the accuracy of detecting embedded nodules. We validate this hypothesis through finite element analysis and the investigation of palpation strategies used by humans during straight unidirectional examination to detect hard nodules inside silicone phantoms and ex-vivo porcine organs. Thus, we conclude that the palpation strategy plays an important role during soft-tissue examination. Our findings allow us, for the first time, to derive palpation behavior guidelines suitable for the design of controllers of palpation robots.