Abstract: An Annotation Platform for the World Wide Web
The one-way flow of information in the World Wide Web restricts the users’ interaction to reading
pages, clicking on links and creating bookmarks. Recent developments in the context of Web 2.0
technology aim at avoiding these restrictions by the use of so-called annotation systems. These
systems provide a way of inserting annotations such as comments or markings into web pages,
thus allowing a more interactive use.
Since in general no client-side writing access to web resources is possible, annotation systems
need to store the annotation data on remote servers and embed them into the original document
by the use of an intermediate system. This can be achieved either in the browser or on a separate
system. However, current approaches have to make a trade-off between an easy way of getting
installed or activated and an intuitive, high-performance user interface. To this end, nowadays still
no widespread annotation system exists.
The objective of this work is to integrate the advantages of both approaches, hence designing
an easy to use and intuitive annotation tool. To this end a proxy server is used, allowing
for inserting JavaScript, HTML and CSS code into arbitrary pages before delivering them to the
browser. This approach firstly avoids the necessity to install client-side software and secondly
offers a way to integrate the annotation functionality directly within the browser, thus being able
to swap computation-intensive tasks to the client side. Additionally an annotation server is provided
from which the data are loaded by using AJAX technology. Yet the user may interact with
arbitrary web pages and enrich them with annotation information without the need for agreement
from the page owner
Abstract: Implicit Annotations and Self-Explanatory Medical Images
In a modern hospital a great deal of information is gathered for each patient. This information
consists of medical images and video files as well as clinical findings and serves as the basis for
diagnosis, treatment and planning of operations. Because of the amount of data it is essential
for the physicians’ work to be able to quickly find and access the relevant information not only
regarding the patient himself but also regarding similar cases to come to the right decision.
As various medical staff are involved in this process meetings and discussions are necessary
to distribute and obtain the relevant information. Here images and video files help to illustrate
important issues.
This project thesis focuses on the development of a system for the storage and processing
of the medical data that are collected during a daily meeting in a hospital and need to be prepared
for usage in the operating theater. These data include various sources such as computer
and positron emission tomographies, magnetic resonance images, endoscopic retrograde cholangiopancreatographies,
medical ultrasonographies as well as text, audio, and laser pointer data.
The sensitivity of this data and, closely connected therewith, the necessity of data protection
in such environments required an intensive examination of the workflow, the technical framework,
and the contents as well as close contact to the physicians who will later be working with the
system, to collect detailed information that was of high relevance for the development process.
The thesis addresses all stages of the development process of a prototype and aims to work out
a specification for a system that could be integrated into a productive environment. The organization
of the thesis is as follows.
First a look is taken at the application domain and the surrounding conditions at the hospital
in order to achieve a deeper understanding not only of the workflow but also of the constraints
that affect the physician’s day-to-day work such as time pressure and data protection policies. The
aim of this analysis was to gain information about how a new system could be introduced into
the existing environment without interrupting the working process and without compromising the
highly security oriented IT landscape.
Based on the results of the contextual inquiry a prototype consisting of two different applications
was designed. The purpose of the first application was to capture all the required information,
convert it and make it accessible in the hospital’s network. The main focus thereby was to automate
and simplify the process in a way such that the expense of time for all people working with
the system was minimized. The second application served the purpose to make the data available
and provide a method to search for specific content. Therefore a web interface based on PHP and
a MySQL database that served as a distribution platform for the content were developed.
During the prototype development process the evaluation took place in an iterative process. On
the one hand surgeons at the hospital evaluated the prototype regarding functionality and content
on the other hand assistants at the Department of Media Informatics at the University of Munich
evaluated the usability of the web interface.
Finally a specification of a possible system based on the results of the contextual analysis and
the cognitions from the evaluation of the prototype will be presented, formulating the requirements
necessary for an extension of the existing system.