Text for Overview Talk
Title: A Collaboration Testbed in Medical Image-Based Examination, Diagnosis,
and Treatment
NSF Grant ECS-94-22701
PI: Terry E Weymouth
Co-PIs: Charles Meyer, Atul Prakash, Thomas Finholt, Ronald Adler, Michael
D. Cohen; University of Michigan.
Post questions and comments to: Terry
E Weymouth (weymouth@umich.edu
).
Contents:
Agenda
- 8:00-8:30 gathering of folks; bagles and coffee
- 8:30-9:00 overview of the project
- 9:00-9:45 Discussion of behavioral observations
- 9:45-10 :00 break
- 10:00-11:00 Discuaaion of Computer Science Framework
- 11:00-11:30 Demos
- 11:30-12:00 Wrap-up
Introduction
In this project we research issues in both the building of collaboration
tools and their use. By interleaving systematic observations of radiologists
and their practice with basic research in collaboration technology and practical
implementation of testbed software, this project is developing new methods
of computer supported collaboration. Details are available in the proposal
abstract.
During the first year of this project, we have developed and evaluated the
first prototype of a session record and replay application. This development
was proceeded and followed by extensive observations of the current practice
of radiologists and primary care physicians. The project integrates systematic
observations of current diagnostic processes with the research and development
of software. See the project summary for the1996
NSF Grand Challenge Workshop for an overview. Here are the details. (Further
details are available at the project web site.)
Behavioral Observations.
In the proposal, we identified the need for a focused study of the actual
user community. Thus, the behavior science component of the Medical Collaboratory
Grant is looking at the potential users of the system from both the clinical
and radiological perspectives.
In the first year, we have completed baseline data collection focusing on
radiological practice and consultations with clinicians. We have gained
access to and have made observations in four different radiological services:
ultrasound, in-patient bone, out-patient bone, and chest. We have made some
limited observations in magnetic resonance imaging.
Additionally, the behavioral science team has identified two satellite clinics
which are potential users of a teleradiology system. One of these is a group
of prinmary care physicians and the other is a group of orthopedic surgeons.
We have also completed baseline data collections at these two sites. The
two sites are interesting because of the differences in their use patterns
of different radiological modalities and the intensity of their use of these
modalities. We are now able to compare our observations with statisitcs
on the radiological use patterns of different satellite clinics extracted
from The University of Michigan Hospital's database.
The data that we have collected is in the form of written notes and videotapes
which document a variety of routine activities as well as exception handling
in these areas. This is the result of approximately 150 hours of observation
with a resulting 80 hours of videotaped information. Data from the Radiology
Department was collected between May and August 1995. Data from the satellite
clinics (primary care and orthopedic surgery) was collected from December
1995 through February 1996.
Data analysis has also begun in terms of coding the data, extracting variables,
and writing up important elements for consideration in designing the first
prototypes of the new system. 70 hours of the videotape has been completely
indexed and analyzed at this point, and the other 10 hours have received
an initial review. Several procedural documents regarding radiological practice
are also now in draft form. Furthermnore, we have developed a codebook and
have extracted elements from the videotaped consultations (n=124) and are
in the process of writing a paper which analyzes these data.
There have been three sessions in which radiologists, behavioral scientists,
and computer scientists have watched the videos together and discussed which
elements must be incorporated into the system. The last one of these focused
on the test of the first prototype. This test was also videotaped, analyzed,
and a list of design considerations was produced.
We created a one-hour summary documentary on radiological practice for the
computer scientists. This identifies the important elements for consideration
in software development and in the analysis of the needs for collaboration.
Finally, we have continued to expand our
bibliogrpahy of articles (>200) relating to the study of radiological
practice and telemedicine applications.
Asynchronous Collaboration.
Our research in asynchronous collaboration focuses on the asynchronous sharing
of workspaces through the use of interactive session objects. Session objects
are intended to capture intra-task content (i.e., how a task was performed
). Our research prototype is being designed to support the replay of a radiologist
application workspace. The asynchronous sharing of this workspace is valuable
to peer radiologists, interns, and clinicians.
A session object encapsulates an interactive session with an application
into a collaboration artifact composed of stored multimedia streams. Our
research directions are on the support of features such as the replay, browse,
annotation, and querying of interactive session objects. We are currently
developing a toolkit ReplayKit to deliver this functionality to the components
of an application workspace. We refer to components integrating our functionality
as replay-aware components.
Our current prototypes extend replay-awareness to the following two components
of an application workspace: visual and audio. The visual component has
two types: input-based (such as the replay of inputs to the interface of
an application workspace) and output-based (such as the replay of display
updates to the display of a workstation). The audio component is needed
to increase the collaboration content of the session. Techniques for the
handling of interactive stored media streams are used on the integration
of these components.
Testbed Development.
For reasons of reuse and portability, we initially focused on Sparc-Solaris2.4
development. We developed an record and replay tool which used:
- Xmx to provide the visual component,
- SPARC audio functions for the audio component, and
- Osiris II for the computational component.
This prototype was evaluated in several
hours of use by a radiologist.
Currently we are developing an extendible architecture that allows the integration
of other replay-aware components, that will provide enhanced Temporal Access
Control to support the features of interactive session objects. We are developing
an objected-oriented version of the testbed in Java. We will begin field
testing this version in August.
Graduate Students.
Following computer science graduate students are working the project(+ indicated
those students that are directly supported by the projects, other students
are supported on closely related projects):
- Nelson Manohar(+): working on asynchronous collaboration using session
record/replay paradigm.
- Amit Mathur(+): working on communication infrastructure for collaborative
systems
- Hyong Shim: working on an object-oriented toolkit for building shared
window applications.
- Gwo-ba Wu: addressing concurrence control issues in collaborative systems.
- Trent Jaeger: addressing security issues in collaborative systems.
- Stephane Cote(+): Collecting, indexing, and analyzing behavioral observations.
Assisting in the development of the codebook and statistical analysis of
data.
- Elizabeth Yakel(+): Site identification and design of observational
protocols. Collecting, indexing and analyzing behavioral observations. Assisting
in the development of the codebook and statistical analysis of data.
.
Papers.
N. R. Manohar and A. Prakash, Tool Coordination and Media Integration for
Asynchronously-Shared Computer-Supported Workspaces.", To be submitted
as a CSE Division Tech. Report, Department of Electrical Engineering and
Computer Science University of Michigan at Ann Arbor, Feb. 1996.
N. R. Manohar and A. Prakash, A Flexible Architecture for Integrating Heterogeneous
Temporal Media on Replayable Workspaces Proc. Third IEEE Int'l Conf on Multimedia
Computing and Systems, Hiroshima, Japan, June 1996, to appear.
N. R. Manohar and A. Prakash, Dealing with Synchronization and Timing Variability
in the Playback of Session Recordings, in preperation.
T. Jaeger and A. Prakash, "Implementation of a Discretionary Access
Control Model for Script-based Systems," Proc. of the 8th IEEE Computer
Security Foundations Workshop, County Kerry, Ireland, June 1995.
N. R. Manohar and A. Prakash, "The Session Capture and Replay Paradigm
for Asynchronous Collaboration," Proceedings of European Conference
on Computer-supported Cooperative Work, Stockholm, Sweden, September 1995.
N. R. Manohar and A. Prakash, Asynchronous Collaboration via The Record
and Replay of Temporal Multimedia Streams, Presented in CSCW'94 Workshop
on Multimedia Infrastructure for Collaboration, Durham, North Carolina,
USA, November 1994. Published in SIGOIS Bulletin, v15, n2, pp32-34, December
1994.
Last modified on March 11, 1996. Terry
E Weymouth.
weymouth@umich.edu