SCH: CyberGut: Cyber-Physical System for Closed-Loop Control of Gastric Motility

Aberrant bioelectrical slow-wave patterns (dysrhythmias) are linked to gastric dysmotility in several significant gastric motility disorders such as gastroparesis, chronic nausea and unexplained vomiting, and functional dyspepsia. Current therapies, including gastric electrical stimulation and pharmaceuticals, are delivered in an open-loop fashion through trial and error, resulting in inconsistent therapeutic effects. No existing tools can identify and deliver optimized electrical stimulation to restore normal stomach motility. This program aims at advancing knowledge and optimizing health treatments in gastric disorders (and more broadly in different organ systems) by developing novel wireless implantable technologies for adaptive, closed-loop recording and stimulation. It overcomes invasiveness of bulky implants, lack of objective patient-in-the-loop feedback in open-loop operation, absence of an interactive model between patient and device for informed therapy decisions, and lack of adaptive intelligent algorithms. This project will develop an intelligent adaptive closed-loop system (CyberGut) that identifies and delivers optimized stimulation for restoring normal gastric motility. Three specific aims include the development of wireless recording/stimulation with millimeter-scale implants; computational organ modeling and machine learning (ML) algorithm developments; and CyberGut system integration and validation. The research tasks will design, develop, and test 1) a network of millimeter-scale devices (called Gastric Seeds), each integrating an application-specific integrated circuit (ASIC) and a magnetoelectric transducer, endoscopically implanted within the stomach submucosal space for fully wireless recording and stimulation functions; 2) a wearable unit for wireless interrogation (powering, communication) of Gastric Seeds through hybrid magnetic-ultrasonic (MagSonic) links, and intelligent identification (through an embedded physics- informed ML model) and delivery of optimized electrical stimulation paradigm to Gastric Seeds; and 3) a computational virtual stomach model within a PC framework that accurately replicates stomach’s physiological/mechanical behavior for generating vast data in training the embedded ML model. RELEVANCE (See instructions): Current therapies (electrical stimulation, pharmaceuticals) for functional gastrointestinal disorders are delivered in an open-loop fashion through trial and error, resulting in inconsistent therapeutic effects. This project enables real-time monitoring, analysis, and control of gastric function through development of an intelligent closed-loop system (CyberGut) that replicates the stomach’s physiological/mechanical behavior.