In the TRANSFORM project, the digitalization of surgeries was the primary focus, centered around the development of the so-called surgical digital twin (SDT).
The concept of a SDT is a complex and specific application of digitalization. The goal is for the SDT to be a perfect virtual representation of the physical world and to create a high-fidelity representation of the OR. For the creation of an SDT the integration of three core elements is needed: the physical object or process and its environment, a digital replication of that environment, and the data and communication links that connect the two. For the alignment of physical and digital entities, there needs to be a common spatio-temporal representation, where data that come from different modalities get fused together.
At the core of the SDT lies a detailed 3D representation of the OR, which serves as the spatial framework for integrating all additional elements. This digital model is then embedded into a VR-system. The OR setting does not only include static elements like surgical lamps and other infrastructural objects, but also includes dynamic entities like medical staff, surgical instruments, and anatomy.
The described VR-System of this project is the first that enables the user to experience a virtual dynamic 3D environment that can be seen as a replica of a real surgery, by encompassing all the different structures. Unlike previous approaches, this project aimed to capture a full segment of a spine surgery instead of only individual anatomical structures or specific surgical interventions.
Capturing this level of complexity requires the integration of various advanced technical devices and sensor systems, which collect data that is then fused with prior domain knowledge to generate an accurate digital model. In this case, five RGB-D cameras brought the surgeon to life in 3D, while a cutting-edge high-end camera captured the anatomical details. Surgical instruments were precisely tracked in real time with an infrared stereo camera, and a laser scanner mapped out the entire operating room, this way seamlessly merging all data into a unified SDT, as shown in the demo video.