The Innovation Centre for Computer Assisted Surgery (ICCAS) promoted by the Federal Ministry for Education and Research (BMBF) is active in the Medical Faculty with several research groups. They carry out interdisciplinary research between the disciplines Surgery, Computer Science, Natural Sciences and Engineering Sciences.
The Research Group “Scientific Methodology” analyzes the proceedings in the operating room and describes these formally by means of ontology and “surgical workflows”. Based on this analysis, software interfaces are developed for the memory and communication of the patient data necessary for surgical intervention. The commonly-used standard in radiology “DICOM” is extended in this sense to “DICOM for surgery”. The formation of modular assistance systems is thereby made possible on the basis, for example, of a Surgical Picture Archiving and Communication System (S-PACS). The concept of “surgical workflow” refers to a description of the flow in the operating room. Research topics are concerned with the development of a recording and analysis methodology for these surgical workflows as well as their usage in order to analyze and assess surgical interventions in detail. The developmental process consists of the following key points: ontological structuring of surgical processes, data acquisition, and data visualisation and analysis. Furthermore, software tools are developed for the individual areas and practical evaluation is carried out for chosen operations in neurosurgery, (distectomies, craniotomies, pituitary adenomas) in ENT-surgery (FESS, MLS, mastoidectomy), in heart surgery (mitral valve replacement, MIDCAB, TECAB), in eye surgery and in interventional radiology.
The Research Group “Surgical PACS and Mechatronics“ is involved primarily with the networking of surgical apparatus in order to offer the surgeon new functionalities. In collaboration with Group 1 “Scientific Methodology”, it is intended that DICOM (Digital Imaging and Communications in Medicine) should be expanded for implementation in the operating room. The goal is to accumulate all patient data within a so-called “world model” and to put said data at the disposal of the surgical apparatus. The surgical workflows are then able to be deployed, from which the requirements are gleaned. Another goal is to specify concrete supplements for the DICOM standard in light of these requirements and to present these to the respective standardization committees.
A prototypical realization of such a networked OP system follows two different approaches that complement one another: (1) the Top-Down Approach analyzes the concrete clinical needs with the aid of clinical workshops, generic workflows and systematic, objective planning procedures- the Quality Function Deployment (QFD). The realizations provided by these efforts flow into the (2) Bottom-Up Approach in which concrete prototypes for special clinical inquiries are developed. Research with prototypes enables direct recognition and solutions for practical and technical problems, as early as in the developmental process.
The summons of Prof. Dr. Dirk Bartz initiated the founding of Research Group 3 “Visual Computing” in December 2006. It is involved with the topics of expanded reality, interaction, visualisation and image analysis. These topics are primarily handled under the category of cognizance (perception).
The following key research points are currently being examined
Ontologies for Surgery
Ontologies represent an efficient opportunity to categorize and formalize our knowledge about the world. In order to efficiently structure ontology for the applied domain “surgery”, we have developed the core ontology COCAS (Core Ontology for CAS), which receives its formal structuring from a Top-Level Ontology. Within COCAS various sub-domains are defined, for which their own domain-specific ontology is allocated. For example, for the structuring of surgical workflows, the ontology “SWOnt” (Surgical Workflow Ontology) was developed. It illustrates all entities, including their prioritization, necessary for describing the surgical processes. Ontologies are, additionally, designed to describe special surgical interventions. FESS-Ont is a good example which is used to be able to formally describe a nasal cavity operation “FESS” (Functional Endoscopic Sinus Surgery).
Surgical Workflows
The surgical workflow is an abstraction of a surgical intervention. Information which is of interest to the user (surgeon, medical technician or health expert) can be ascertained from reality. In order to enable the use of surgical workflows (e.g. for purposes of analysis), one requires technical and formal descriptions of the surgical intervention. This brings about the need for structured acquisition of information from the inter-operative surgical reality.
In order to support the recording process, a software-architecture for the protocolling of surgical interventions was developed at ICCAS- the ICCAS Surgical Workflow-Software. With the aid of this program it is possible, for the first time, to gain detailed information regarding the sequential run of an operation, and to assess this according to clinical and technical formulation of questions.
The analysis of surgical workflows had previously been used for the definition of Use Cases for the preparation of DICOM-objects. There exist as well experiences with the specification of virtual surgical simulators on the basis of surgical workflows.
This methodological approach has been used up to this point for the protocolling of more than 200 interventions. This methodology has also been used across disciplines in the fields of neurosurgery, ENT-surgery, cardiovascular surgery, eye surgery and interventional radiology application.
Modular Assistance Systems for Surgery / DICOM for Surgery
DICOM represents an established standard for the storage and transport of image-data in radiology. ICCAS is occupied, within the framework of the DICOM Working Group 24, with expansion of this standard for use in the operating room. In this context an initial design was devised for the recording and storage of surface networks which flow into the standard. The efforts in the area of “DICOM for Surgery“ comprise the software-interface basis for the “Surgical PACS“ (see below).
Clinical applications on the basis of developed specifications are realized in prototypical scenarios. For example, image-processing problems and modelling problems were solved in various projects.
In addition to segmentation of blood-vessel branches in angiographs and specification of the working space for navigated sinus operations, methods of classification for various types of tissue were developed from image data. The modelling techniques are implemented in a patient model, which puts all collected informational data regarding the patient at the surgeon’s disposal in an optimal manner.
Virtual Reality and Training Systems
Techniques for virtual reality were evaluated regarding their applicability in a clinical routine. In this context, the usability of three-dimensional representations was evaluated in numerous studies in ENT-surgery and neurosurgery.
Three different virtual training systems were individually evaluated with force-feedback for a sinus OP, mastoidectomy and ventriculocisternostomy OP (Torkildsen’s operation). Various parameters, such as haptic feedback, can be incorporated, and learning curves can be constructed. The differentiated effect upon both experienced and inexperienced surgeons is also evaluated here.
Completely new approaches are being evaluated in the area of patient-specific creation of models. A sub-project consists of the application of image-processing methods, the simulation of neurosurgery and ENT-surgery, and the planning of intervention for accident surgery.
Open Network Infrastructure for the Operating Room of the Future
(Surgical PACS)
The following interventions were chosen for the future application of a surgical PACS system:
- Neuro-Navigation with the Aid of Intra-Operative 3D Ultra-Sound
Intra-operative image reproducing procedures, such as ultra-sound, can be directly implemented during an operation and improve orientation at the operation site as well as control over the resection sequence. The combination of conventional image reproduction (MRT) and intra-operative 3D ultra-sound (3D-IUS) communicates a plethora of supplementary information to the surgeon. With this information applied to neurosurgery, the positional relationship of organs to one another, and among other things, the shifting of brain tissue or traces of tumour remnants can be spatially visualized. As compared to other intra-operative image reproduction procedures (e.g. IMRT), the 3D-IUS-System represents a highly compact and flexible instrument for the operator. By means of joint developments in collaboration between the Firm Localite and ICCAS, it was possible, as early as 2006, to improve the 3D ultra-sound system in essential aspects, such as by the shortening of the duration of the instrument calibration. In one aspect of these joint efforts, the concept of open and unified apparatus interfaces for communication between various system components has already been tested.
- Expanded Reality in Heart Surgery
In cooperation with the Heart Centre of the Universität Leipzig, a system is being developed that makes it possible to superimpose model images onto the real endoscopic images, in order to simplify the identification of various anatomical structures. During minimal-invasive surgery on the heart, the field of vision is limited, making the viewer’s orientation more difficult. Landmarks which provide orientation and simplify identification of the target area are often times not visible due to the endoscopic access. Specifically, when utilizing the daVinci-Telemanipulator for heart-surgery interventions, finding the target vessel for the bypass apparatus can be impeded or even impossible.
Expanded Reality is a technology which can be implemented for such situations as an aid in that a virtual 3-D model is superimposed upon a camera image in real time. The angiography has always been the diagnostic state-of-the-art standard for coronary heart disease. A virtual 3-D model, unique to a particular patient, can be created from the angiographic data. This is superimposed upon the real endoscopic image during the operation and provides both orientation and navigation when searching for the target vessel for the bypass. Within the project, software and equipment technology for clinical deployment in bypass operations is being developed, with which an expanded reality in the form of a virtual 3-D model is utilized.
Navigated Control – Performance-Guided Assistance Systems for Surgery
The assistance system for functional endoscopic sinus surgery (FESS – Control) was evaluated at the Clinic and Poly-Clinic for Ears, Nose and Throat Medical Science. The manually-guided instrument turns itself off when its position moves outside the working space defined via pre-operational segmentation. This principle is called Navigated Control and is integrated into the various applications at the Institut für Mikro- und Medizintechnik (MIMED) of The Technical University of Munich. Within the framework of this study, 10 patients have been treated so far- all without complications. Clinical and ergonomic parameters have shown improved quality of both the information in the navigation data and consequent changes in the surgical procedures. Furthermore, mental relief for the surgeon was also substantiated.
Based on the positive results with the FESS-Control System, the same principle should be applied to interventions in the petrous bone (mastoid). A mastoidectomy represents a complex and risk-laden operation, and Navigated Control should be able to aid this procedure by switching off a bone drill that manoeuvres outside the working space (e.g. in regions with blood vessels or facial nerves). This solution combines the advantages of the robot-supported systems and the pure Navigation Systems. The surgeon guides the drill. The Navigation System makes sure that the instrument is switched off if it moves into an improper position. The pre-clinic evaluation of Mastoid-Control was completed the end of 2006. A clinical multi-centre study with the Klinik und Poliklinik für HNO-Heilkunde of the Universität Leipzig and the Medical College of Hannover will be initiated in 2007.
First of all, the system was evaluated with an electronic Phantom (ElePhant). For this purpose, an anatomical model with 3D-printer technology was produced. There is a micro-controller within the anatomical model which assumes the communication between model and PC. The doctor can simulate the intervention with the real surgical instrument. Any damage to high-risk structures is objectively recorded during the simulation. Evaluation parameters, such as time and quantity of the damaged high-risk structures, for example, are automatically stored.
A third area of application for Navigated Control is a navigated and guided drill for spinal column surgery. The surgeon segments a working space during the operation. In this case the navigated drill is also automatically activated, as soon as it is placed within the working space, and stops automatically when it leaves the working space. For pre-clinic evaluation, an anatomical spinal column model (ElePhant) with a 3D- printer was also produced. The dura mater (meninx), as a high-risk structure, was built of conductible material. And then the model was connected to an electrical circuit. If the surgeon damages the high-risk structure during the simulation, this damage is objectively detected by the computer. The segmentation precision of the intra-operative X-ray (C-Arm Image Intensifier) was evaluated first of all. It was shown that the segmentation of the working area with the aid of the 3D C-Arm could ensue in a sufficiently precise manner. In the next study, the precision of the drill when used on the anatomical model will be analyzed.
Clinical Evaluation
As regards a general evaluation of computer-assisted surgical instruments, the technical quality of systems has, for the most part, been ascertained, while the impact upon patients and the ergonomics for the surgeon have only been analyzed minimally. In order to enable international and inter-institutional comparability, guidelines were created based on six levels of evaluation – from technical to clinical, ethical, sociological and legal criteria. Based on these levels and the accepted evaluation guidelines, the systematic study-protocols and recording questionnaires of other medical subject areas are, step by step, adjusted to computer-assisted surgery.
Home 
