Medical records are being digitized in order to improve patient care, and the integrated digitization of all components of the medical record will also help reduce the number of errors and reduce health care costs. Images form part of the medical record. In some specialty areas such as radiology, the use of digital imaging is widespread. However, in other specialties digital images are less frequently used. One example is anatomic (surgical) pathology in which a microscope is used to view tissue mounted on a glass slide to obtain an image, a practice that has not changed much over decades. Pathologists have been slow to transition from glass slides to digital imaging for reasons which relate to image acquisition and interpretation. High-resolution digital images are necessary for accurate diagnosis, and acquisition of these images has been slow and expensive. Advances in computer technology have increased the speed of acquisition and reduced costs such that slide digitizing systems are now available from major microscope manufacturers. For digital imaging to become widely accepted in pathology there is still need for innovation to improve performance at all stages of the imaging process including acquisition, user interfaces, display, analysis, access and long-term storage. Here we propose development of a freely available open source system that can operate on multiple different computing platforms (Windows, Mac OSX, Linux) and process slide image data for integration of digital pathology into hospital-wide clinical information systems. The client/server software will manage collection of the image data and create large high resolution images that can be streamed over standard networks for display and remote viewing. Features of the software will include the ability to annotate images as well as support three-dimensional multi-channel datasets. This will enable pathologists to rapidly step through images of successive sections from a tissue block and facilitate three-dimensional analysis of the lesion biopsied. As pathologists require an efficient and fast workflow and may have limited computer expertise, our goal will be to have software that is as fast and easy to use as a microscope- the standard by which pathologists will judge the usefulness of the tool. We will develop and evaluate several client applications, one based on a standard desktop workstation interface paradigm, another deployed within a web browser, and finally a third in a multi-touch environment supporting direct interaction with images. In addition the software will have a plug-in interface (API) that enables development of image processing modules as well as new client interfaces. Ultimately the open source system will have advantages for training, education, remote/collaborative viewing, and research.

Public Health Relevance

Pathology has a storied history in the medical profession and is vital to the diagnosis and treatment of disease. Digital pathology has the potential to advance the efficiency and accuracy of the field, providing new tools for interaction, analysis, collaboration and education.

Agency
National Institute of Health (NIH)
Institute
National Institute of Biomedical Imaging and Bioengineering (NIBIB)
Type
Small Business Innovation Research Grants (SBIR) - Phase I (R43)
Project #
1R43EB013518-01A1
Application #
8199429
Study Section
Special Emphasis Panel (ZRG1-SBIB-Q (80))
Program Officer
Pai, Vinay Manjunath
Project Start
2011-07-15
Project End
2012-06-30
Budget Start
2011-07-15
Budget End
2012-06-30
Support Year
1
Fiscal Year
2011
Total Cost
$149,353
Indirect Cost
Name
Kitware, Inc.
Department
Type
DUNS #
010926207
City
Clifton Park
State
NY
Country
United States
Zip Code
12065