Research Projects

 


KISMET: Knowledge & Information Sharing in Medical Expert Teams (2015-2018)



In the field of endoscopic surgery, modern information technology brought revolutionary improvements in the last decade. Physicians are enabled to perform minimally invasive surgeries with the help of high resolution cameras, to follow the progress of the surgery on large, high definition (HD)  video screens and to record and store the videos of the interventions. These techniques are on the highest level of state-of- the-art technology. On the other hand, the further destiny of the recorded endoscopic videos falls into a technological gap. Tools for further exploration, such as search, interaction and sharing are poor or non-existing at all. The main goal of the KISMET project is to fill this gap, at least to a certain extent, with novel and competitive tools.

We concentrate on a specific area, on endoscopy in gynecology, in particular in endometriosis, relying on the world-wide admitted practices of the LKH Villach in this area. We apply basic and applied research methods to investigate the current situation in the hospital LKH Villach, identify problems in the workflows related to multimedia content, develop possible improvements based on a self-organizing cooperative system and evaluate this system in day-to-day usage scenarios. The three main research questions of the project are (i) how to gain expert knowledge from physicians in a convenient way, without imposing additional workload on them, (ii) how to adapt state-of-the-art multimedia content indexing methods to the specific characteristics of medical multimedia data based on expert knowledge and (iii) how to support sharing of data and metadata among the experts (enabling an enhanced communication with patients as well).
Funded by:

  • KARL STORZ GmbH & Co KG
  • The European Regional Development Fund and the Carinthian Economic Promotion Fund (KWF), supported by Lakeside Labs GmbH, Klagenfurt, Austria

Duration:

  • October 2015 – September¬†2018

 Project leader:

  • Assoc. Prof. Dipl.-Ing. Dr. Klaus Sch√∂ffmann

Employees:

 

Juergen Primus Stefan Petscharnig Andreas Leibetseder
Dipl.-Ing. Dr. Manfred J√ľrgen Primus Dipl.-Ing. Stefan Petscharnig Dipl.-Ing. Andreas Leibetseder
Post-doc researcher PhD student PhD student

 

 


 EndoVIP-2: Endoscopic Video Processing 2 (2015-2018)


In recent years, it became common to record video footage of endoscopic surgeries. This leads to large video archives that are very hard to manage. They often contain a considerable portion of completely irrelevant scenes which waste storage capacity and hamper an efficient retrieval of relevant scenes. Common lossy video compression methods have already reached a very high performance level and are about to reach their natural limit. However, their compression efficiency is still insufficient for certain domains where a plethora of video data should be archived. One example is the domain of medical endoscopy where entire surgeries are recorded for documentation with high visual quality because it is not possible to anticipate if and when exceptional situations may happen. Eventually, only a fraction of the video footage is of actual relevance for post-operative review. Therefore, in this project we investigate domain-specific compression methods for endoscopic videos as well as special video interaction tools to facilitate content-based search in archives of endoscopic video data.

Funded by:

  • KARL STORZ GmbH & Co KG

Duration:

  • October 2015 – April 2018

Project leaders:

  • Assoc. Prof. Dipl.-Ing. Dr. Klaus Sch√∂ffmann
  • Univ.-Prof. Dipl.-Ing. Dr. Laszlo B√∂sz√∂rmenyi

Employees:

Bernd Muenzer
Mag. Dipl.-Ing. Dr.¬†Bernd M√ľnzer
Post-doc researcher

More information…

 


SQUASH: Surgical Quality Assessment in Gynecologic Laparoscopy (2016)


Laparoscopic surgery requires specific psychomotor skills of the operating surgeon, which are difficult to learn and teach, and typically result in prolonged learning curves. These psychomotor skills, though, have direct impact on the performance of the surgery, especially in fields with complex operation techniques. In order to assess surgical quality objectively, medical experts have proposed to inspect and analyze the unedited video footage from endoscopic surgeries for the occurrence of technical errors. However, manual inspection of video recordings is a time-consuming and error-prone task, which is currently performed with a common video player, external checklists, and manual notes. In this research project, we want to find out how we can efficiently improve surgical quality assessment through computer support. We want to investigate fundamental aspects of computer-aided surgical quality assessment, in particular in the field of gynecologic laparoscopy.

Funded by:

  • Medical University of Vienna / AKH Wien

Duration:

  • January 2016 – December 2016

Project leader:

  • Assoc. Prof. Dipl.-Ing. Dr. Klaus Sch√∂ffmann

Employees:

 

Marco Hudelist Sabrina Kletz
Dipl.-Ing. Dr. Marco Andrea Hudelist Sabrina Kletz
Post-doctoral researcher Student assistant

 


NGVB: Next-Generation Video Browsing (2012-2016)


Video browsing is known as the interactive process of quickly navigating through the content of a video in an explorative fashion. In difference to video retrieval the purpose of video browsing is manifold: (1) get a quick overview of the content of a video and its structure, (2) find out where potentially interesting segments are located, and (3) interactively search for a specific part of content that cannot be found through video retrieval applications. Although many video browsing approaches have been proposed over the past years, there is great potential to significantly improve existing tools by exploiting features of modern mobile devices. In the Next-Generation Video Browsing project we want to translate our results from previous research on video Screenshot9browsing and 3D search interfaces to this interesting and challenging new domain and answer several new research questions. We focus on video browsing with modern tablet computers that provide high computing power, new interaction features, and communication functions that enable completely new use scenarios for video browsing.

Funded by:

  • The Federal Ministry for Transport, Innovation and Technology (bmvit) and the Austrian Science Fund (FWF): TRP 273-N15
  • The European Regional Development Fund and the Carinthian Economic Promotion Fund (KWF), supported by Lakeside Labs GmbH, Klagenfurt, Austria

Duration:

  • August¬†2012 – January¬†2016

 Project leaders:

  • Ass.-Prof. Dipl.-Ing. Dr. Klaus Sch√∂ffmann
  • Univ.-Prof. Dipl.-Ing. Dr. Laszlo B√∂sz√∂rmenyi

Employees:

Marco Hudelist Claudiu Cobarzan
Dipl.-Ing. Marco Andrea Hudelist Dipl.-Ing. Dr. Claudiu Cobarzan
PhD student Post-doc researcher

More information on the project website of NGVB.

 


CODE-MM: Community of Domain Experts in Medical Multimedia (2012-2015)


Collaborative knowledge in big communities (known as Wisdom of the crowds) and self-organization of huge amount of networked participants have shown a lot of potential in diverse fields of computer science and information systems. Still, there are many scenarios, in particular in highly professional applications, where a critical mass of users (experts) cannot be found in order to safely create this knowledge. In several scenarios a top down model  providing generally agreed domain knowledge can be used as common ground for communication. However, for other scenarios, diversity in terms of interest, motivation, information need, and many other characteristics prevent top down approaches, for which an à priori consensus between participating members must be found. Within the project at hand we focus on small and medium sized groups of people, who (i) have common interests but are too diverse for applying a commonly agreed model and (ii) who are bound to use multimedia systems to satisfy their needs. Further, we consider these people being experts with intense domain knowledge, such as medical doctors.

In our project we plan to investigate (i) how communities of domain experts can be supported by multimedia information systems and (ii) how to leverage self-organization within those groups to a new level by novel and innovative multimedia services. With the help of our project partner, the KARL STORZ GmbH & Co KG, we will employ our research results to a real world scenario: a multimedia information system for endoscopic videos supporting surgeons spread all over the world. While content storage can be regarded as a solved problem, intelligent and specialized content annotation, retrieval, and analysis methods are totally missing in this domain. Tools for exchanging information  among the domain experts are missing as well.

Funded by:

  • The European Regional Development Fund and the Carinthian Economic Promotion Fund (KWF) under grant KWF-20214 22573 33955, supported by Lakeside Labs GmbH, Klagenfurt, Austria
  • KARL STORZ GmbH & Co KG

Duration:

  • February¬†2012 – June¬†2015

 Project leaders:

  • Ass.-Prof. Dipl.-Ing. Dr. Klaus Sch√∂ffmann
  • Univ.-Prof. Dipl.-Ing. Dr. Laszlo B√∂sz√∂rmenyi
  • Ass.-Prof. Dipl.-Ing. Dr. Mathias Lux

Employees:

  • Dipl.-Ing. Dr. Manfred del Fabro (Post-doc researcher)
  • Dipl.-Ing. Mario Guggenberger (PhD student)
  • Dipl.-Ing. J√ľrgen Manfred Primus (PhD student)

 


EndoVIP-1: Endoscopic Video Processing (2012-2015)


In recent years, digital images and videos of endoscopic screenings and surgical operations are recorded and archived by physicians with increasing frequency. The reason is that interventions can be documented in a much more expressive and informative way than with textual descriptions only. But an extensive, naive usage of endoscopic videos leads to huge amounts of video data that can easily exceed the storage capacity of hospital operators. Moreover, it is a very cumbersome task to retrieve specific information from such a large data set without efficient tools that are tailored to this very specific domain. Therefore, the purpose of EndoVIP is to investigate how this video data can be significantly reduced without losing any important information. This involves the development of efficient encoding strategies as well as the automatic identification of irrelevant scenes by means of content-based analysis methods. If irrelevant scenes are omitted and only the actually relevant scenes remain, also the browsing and retrieval performance can be improved considerably. In this context, novel mechanisms for time-efficient retrospective viewing and retrieval of specific relevant scenes will also be investigated.

Funded by:

  • KARL STORZ GmbH & Co KG

Duration:

  • May¬†2012 – April¬†2015

Project leaders:

  • Ass.-Prof. Dipl.-Ing. Dr. Klaus Sch√∂ffmann
  • Univ.-Prof. Dipl.-Ing. Dr. Laszlo B√∂sz√∂rmenyi

Employees:

  • Mag. Dipl.-Ing. Bernd M√ľnzer (PhD student)