Cyber-physical Twins of Human Organs

Overview

Surgeons, like airline pilots, need realistic environments in which to practise complex procedures. However, existing surgical simulators and simple substitutes often fail to reproduce the anatomy, texture, and mechanical behaviour of human organs.

Led by Tian Qiu, this project developed cyber-physical organ phantoms in collaboration with three medical departments at the University Hospitals of Tübingen and Freiburg. The team created and studied six realistic models: the kidney, bladder, prostate, liver, vagina, and brain. These models combine physical replicas of organs with sensing and digital technologies, making it possible to collect biomedical data that would be difficult or impossible to obtain during real surgery.

Key results

• Developed organ models using 3D printing, soft-material moulding, and hydrogels designed to reproduce realistic anatomy and tissue mechanics.
• Created a brain phantom made from a transparent composite hydrogel, enabling high-resolution imaging while preserving tissue-like mechanical properties.
• Integrated optical imaging, ultrasound, electrical sensing, and augmented-reality tools into the physical models.
• Developed systems for the quantitative assessment of surgical skills. Computer-based analysis can evaluate factors such as target detection, instrument movement, visualisation, and procedural efficiency.
• Demonstrated that automated video analysis could distinguish between experienced surgeons and trainees during simulated endoscopic procedures.
• Developed a hybrid surgical simulator in which a physical urinary-tract model is combined with an AI-based digital system that detects lesions and provides quantitative feedback.
• Enabled clinicians to practise procedures and evaluate surgical instruments in realistic, repeatable, and safe environments.

Why it matters

The project shows how realistic organ phantoms can function as both surgical-training platforms and experimental biomedical systems. They allow researchers to measure surgical performance, test instruments, and generate data for computer-assisted training without relying exclusively on animal tissue or human patients.

The work could support future applications in surgical education, medical-device development, and AI-assisted assessment.