Machine learning-enhanced soft robotic system inspired by rectal functions to investigate fecal incontinence

Abstract

Fecal incontinence (FI), which can arise from various pathogenic mechanisms, has attracted considerable attention worldwide. Despite its importance, the reproduction of the defecatory system to study the mechanisms of FI remains limited, largely because of social stigma and being considered inappropriate. Inspired by the rectum’s functionalities, we developed a soft robotic system that includes a power supply, pressure sensors, data acquisition systems, a flushing mechanism, stages, and a rectal module. Specifically, the innovative soft rectal module includes actuators inspired by sphincter muscles, both soft and rigid covers, and a soft rectum mold. The rectal mold, which was fabricated from materials that mimic human rectal tissue, was produced using a mold replication fabrication method. Both the soft and rigid components of the mold were created using three-dimensional (3D) printing technology. In addition, the sphincter muscle-inspired actuators featured double-layer pouch structures that were modeled and optimized based on multilayer perceptron methods to obtain a high contraction ratio (100%), generate high pressure (9.8 kPa), and have a short recovery time (3 s). Upon assembly, this defecation robot could smoothly expel liquid feces, perform controlled solid fecal cutting, and defecate extremely solid long feces, thus closely replicating the functions of the human rectum and anal canal. This defecation robot has the potential to facilitate human understanding of the complex defecation system and contribute to the development of improved quality-of-life devices related to defecation.