Open Source Telepathology Project Released

iPath is a recently released open source telepathology project. iPath is intended to be a modular, easy to use, web-based telepathology solution with no specific software and hardware needs. Currently two modules are released and running: a) At the core, the telepathology database for case discussion and second opinion consultations and b) a telemicropscopy module for real time microscope sharing over internet.

iPath – Internet Pathology Suite

iPath originated from our frustrations with existing internet-telemicroscopy solution which usually ended within our hospitals firewalls. After finding a technology to do remote microscopy over internet connections which are secured with firewalls, we needed a versatile database which allowed us to store images generated in telemicroscopy sessions.

While developing such a database we became aware that the main benefit of telepathology is its value for second opinion consultations, a field generally regarded as the most beneficial part of telepathology (cf. Mairingen 2001 in ACP 21(3,4)).

Once our work reached a certain stability, we decided to release our Software collection as an open source project to allow others to participate and use it. Currently there are two modules released:

Telepathology Database: A database with a web based userinterface (PHP, PostgreSQL and Apache) that allows a group of pathologists to discuss difficult cases online.
real-time Telemicrosopy: A small piece of software that allows to connect a conventional or a motorized microscope to an iPath-Server from where it can be accessed and controlled over a Java-Applet from one or more remote pathologists.

People interested in our telepathology database are welcome to visit our
test
server
.

Or visit the project page at: ipath.sourceforge.net, where the software is released.

PS: contributions and comments are highly welcome.

Semi-Sentient Medical Software?

Los Angeles Times has an article about an artificial intelligence project called ‘Cyc’ for encyclopedia which has been in existence for approximately 17 years with the goal of overcoming the ‘common sense’ problem with artificial intelligence. Namely that computers can do logic quite well, they simply lack common sense knowledge necessary for intelligence. For example, a computer has to be told that red spots on an automobile are unlikely to be caused by measles, but are a possibility in humans. However, entering common sense knowledge is a daunting challenge which now appears to be nearing a usable level with 1.4 million assertions entered into Cyc’s knowledge base. Since its inception, medical applications have been expected from this project and the prospect of semi-sentient medical record software is fascinating. The full knowledge base is only available to licensees, but a small open source version with the same inference engine Updated 6/27/01, 12:08 CST: is scheduled for release ‘late summer 2001’ under the GNU Lesser General Public License (LGPL) according to the OpenCyc website.

JoelOnSoftware: Let Me Go Back!

JoelOnSoftware.com has an article by Joel Sposky on how to succeed with software in a market already dominated by a competitor. Suggestions: admit that it is dominated by a competitor, find ways for customers to be able to switch, and lower the barriers to entry for your product. Sposky can speak authoritatively because he was a developer on products you may have heard of like Microsoft Excel and some original design work that became Visual Basic for Applications.

RedHat Achieves 1st Quarter Profit

RedHat, Inc. maker of the most popular distribution of Linux is reporting its first-ever profit which occurred during the 1st quarter 2001: ‘…”In an increasingly difficult IT environment, Red Hat delivered a profit and generated positive cash flows for the first time,” commented Matthew Szulik, President and CEO of Red Hat. “Companies continue to migrate their platforms to Red Hat Linux and to adopt Open Source solutions for mission critical applications.” One can argue that Microsoft has been profiting from open source software for years since the Wall Street Journal reports that despite Microsoft’s denials, it has been using open source software in its ‘major products’ for years.

MozillaQuest Magazine: Netscape 6.1 PR 1 Browser Released

Looks like the wait for a competitive open source web browser is nearing its end. MozillaQuest Magazine has an article announcing that Netscape 6.1 Pre Release version 1 is now available and is based on the Mozilla 0.9.1 code base: ‘…Simply put, the Netscape Division grabs the Mozilla source code, adds its own and third party products such as Net2Phone and AIM, and then puts its own Netscape branding on the added-too and customized Mozilla-base. The Netscape Division also adds its portal and bookmark items to the Mozilla-base while customizing the underlying Mozilla browser suite. Stir all that together and voila, that’s the Netscape 6.x browser suite. All told, the Mozilla browser suite upon which Netscape 6.1 is based is a nice browser…’ The article has links to download sites for Linux, Mac and Windows.

December 2001 Convergence

Some of the subscribers to my newsletter are colleagues and associates in health care that I meet with socially but are only peripherally aware of the goings-on in open source software. Many of them have noted how GNU-Linux (usually referred to as just Linux) was supposed to take over the world one year ago and wonder: what happened, where is it? To which I reply: December 2001. Here’s why.

To the casual observer of the computing scene, not much has changed: Windows is still entrenched, MS Office is the de-facto standard and seems to have a continued raft of products such as Windows XP. The uproar in the media for Linux and the valuations of Linux software companies such as RedHat have both subsided. However, to conclude that nothing has changed would be a mistake.

The fundamentals of Linux are still there: it is still a compelling software platform both economically and technically particularly as an Internet server where it is a force to be reckoned with. What hasn’t happened is a takeover of the desktop. There is a simple reason for this: it isn’t ready.

I qualify this with: all of the distributions aren’t ready for the average user. I’ve been using it on my desktop exclusively for over a year now and find it superior in many ways. My wife who is as non-technical as can be uses it also for school and documents using StarOffice 5.2. Then again, she has me to admin her machine. There are some end-user experience issues which keep Linux out of the reach of the masses: 1) Installation of video and sound as well as other installation difficulties remain an issue. 2) Anti-aliased fonts are not widely available through all the distributions. 3) A browser with the familiar Netscape name is not currently competitive. 4) Some application software is either a) not ready, b) not as good as applications such as MS-Office, or c) ready and superior to its Windows equilvalents (see my recent article on scanning) but requires more effort and knowledge on the part of the user to find and use.

Given the current rate of development and release of Linux and other open source applications, December 2001 appears to be the month that a usable convergence of many of the above open source technologies will occur. At that time Linux distributions are likely to have most of the necessary software required in a form that can compete in all areas. Interestingly enough, open source medical computing will also likely have its first viable offerings at the same time.

The most un-certain part of this analysis is a more competitive alternative to MS-Office than last years release of front-runner StarOffice release 5.2 which is growing stale. OpenOffice appears to be moving a long, and is quietly and un-officially predicting a 3rd quarter delivery of OpenOffice which will possibly occur simultaneously with Sun Microsystem’s commercial version of OpenOffice, StarOffice 6.0. Delivery of a competitive version of OpenOffice and StarOffice 6.0 will be crucial to acceptance of Linux on the desktop.

I chat up Linux a lot publicly and privately as editor of Linux Medical News. What I haven’t done is recommend Linux to the average user. All that is likely to change in December 2001.

Medical Record Exit Strategies

John Faughnan has written a thought provoking piece on something that few organizations think about: What happens if an organization wants to switch electronic medical record (EMR) vendors? ‘…I think this is an interesting question. There are many reasons a practice might want to change EMR vendors. A practice might be acquired, or it might be sold. It could split or merge. It could change character (FP to multispecialty). A vendor might raise rates, or go out of business, or change markets, or leave the market. A product could be suited to a small practice but not a large one, to a single site but not multiple sites, etc…’ Lots of good links in this one.

Structured Analysis/Design Methodologies & HIPAA

Here is one of the first articles I’ve seen that explores the approach of applying structured analysis methodologies such as Unified Modeling Language (UML), definition here, to the HIPAA challenge.

Though I’m an advocate of both conventional (Gane/Sarson) and more current UML methodologies with respect to system/software development and business process re-engineering, ten years of professional practice as a business systems analyst has demonstrated that the majority of organizations will not pay for the professionals and the Computer Aided Software Engineering (CASE) tools needed to support modern-day structured analysis and development.

Those organizations that do fund a structured approach like UML have a tough-time suffering the time frames needed to gather requirements, analyze needs and document the results, before beginning to develop and implement a new design. This is true even for executives and managers who understand the long-term benefits of this approach over conventional analysis and development methodologies.

Extreme programming/data modeling, and even a condensed UML approach (Extreme UML) are in the works by some advocates, in an attempt to shorten the lifecycle of a structured analysis and design project. If these efforts are successful, one expected result would be that more organizations would adopt and fund a structured analysis and development approach, as opposed to hacking away endlessly trying to bring an organization into compliance with HIPAA.

it-costs-too-much-compared-to-what

Medscape (free login required) has a conference report on how technology is changing medical education, particularly via PDA’s. What is notably missing is actual patient records at the point of care. ‘…Currently…most medical students do not use their PDAs for patient-encounter information. Deleshia Kinney, a first-year medical student from Southern Illinois University, noted, “Right now I use my PDA for references, PBL questions, and research. I know that they will be useful for clinical encounters and annotation, once the issues of confidentiality, security, and ease of updating software have been addressed.”