Surgical workflow recovery is a crucial step towards the development of intelligent support systems in surgical environments. The objective of the project is to create a system which is able to recognize automatically the current steps of a surgical laparoscopic procedure using a set of signals recorded from the OR. The project adresses several issues such as the simultaneous recordings of various signals within the OR, the design of methods and algorithms for processing and interpreting the information, and finally the development of a convenient user interface to display context sensitive information inside the OR. The current clinical focus is on laparoscopic cholecystectomies but the concepts developed in the project also apply to laparoscopic surgeries of other kinds.
Providing feedback to trainees is one of the most important issues to support learning. This project researches how to give quantitative and visual feedback on the performance of a student when training on simulators or phantoms. Statistical analysis and probabilistic models are used to compare the performance of students and experts. Augmented Reality and video is used to give visual feedback like a synchronized replay of the student’s and the expert’s performance.
The project Augmented Reality Supported Patient Education and Consultation (Augmented Reality unterstützte Operationsaufklärung) aims at developing Augmented Reality (AR) supported communication tools for patient education. The development of the targeted systems involves disciplines reaching from image registration, human computer interaction and in-situ visualization to instructional design and perceptual psychology. As a primary clinical application, we determined breast reconstruction in plastic surgery.
We are simulating Ultrasound images from CT volumes by assuming a correlation between Hounsfield units and acoustic impedance. Rays are cast trough the CT volume to simulate the US image formation. Simulated Ultrasound from CT can be used for CT/Ultrasound registration by comparing a real US image to simulated ones. Another application of simulated US is training.
In this project, we aim at discovering automatically the workflow of percutaneous vertebroplasty. The medical framework is quite different from a parallel project , where we analyze laparoscopic surgeries. Contrary to cholecystectomies where much information is provided by the surgical tools and by the endoscopic video, in vertebroplasties and kyphoplasties, we believe that the body and hand movement of the surgeon give a key insight into the surgical activity. Surgical movements like hammering of the trocar into the vertebra or the stirring of cement compounds are indicative of the current workflow phase. The objectives of this project are to acquire the workflow related signals using accelerometers, processing the raw signals and detecting recurrent patterns in order to objectively identify the low-level and high-level workflow of the procedure.
Research Paper / IDP: Economic Assessment of Using Augmented Reality in the Construction of Power Plants (At the chair for information systems and in cooperation with Siemens CT)