We believe it is important to have the best possible user experience for participants filling out intake forms in Athena. For this reason, we are developing a *native* iPad application that implements the Athena intake questionnaire and submits the responses to the Athena HQS database running on the Force.com platform.  

The Athena project (see below) is implementing a electronic capture of patient reported information (intake forms).  Due to a technology gap regarding the electronic capture/submission of patient intake forms, we developed a configurable, multi-questionnaire capable system that Athena sites will use to enable electronic intake forms in the mammography clinic waiting room, or prior to the mammography appointment. We elected to move towards using a "platform as a service" development model (PaaS) and have worked with the Salesforce Foundation in implementing the Athena HQS system as a Force.com platform application available to all the UC Athena participating sites. 

MOJO - Medically Oriented Java-based Ontologically Aware System

A project of one of my graduate students, Bill Reidl, the system employs a number of natural language processing techniques to undertake concept identification.  Given concept-based auto-coding of healthcare information is highly dependent on correctly spelled content, MOJO includes a spell checking module. Both the concept identification and spell checking functionality have been implemented as REST services. 

The Athena Breast Health Network is a large-scale project designed to change breast cancer detection and care by merging research, technology, and novel healthcare delivery models to optimize the translation of new research findings into patient care. Athena requires a large-scale informatics infrastructure to manage the 150,000 record registry, which includes anonymized patient information, radiology/pathology images, and genomic data. I am currently the Chief Medical Information Officer (CMIO) for the Athena project and work with informatics/HIT groups at each campus to manage Athena related data.  More about Athena.

A collaboration with Dr. Laura Esserman (UCSF Breast Cancer Center of Excellence) and the NCI Center for Bioinformatics (CBIIT) to build an integrated informatics architecture that supports adaptive clinical trials. The system is used to support the ongoing I-SPY 2 Breast Cancer Trial.  TRANSCEND uses several caBIG components as well as an open source web-based secure clinical information system for data capture at the point of care. 

In 2003, I was very fortunate to have been approached by the California Department of Public Health about working with them to development of a large-scale, secure, web-based, death certificate registration system.  We assembled teams in the Univ and the Dept. of Public Health and began building a system that has become one of the largest and most feature-rich death registration systems in the US.  CA-EDRS was first deployed Jan 1, 2005 after a 13 month development.  The system today handles 99.8% of all death certifications in California. We pioneered the use of fax and voice attestation in vital record systems.  This approach is now being used by a number of other states. 

--------------------------------------------------------------------------------------------------------------------------------------------------------

jTerm (1999 - 2001)

An open-source java-based terminology server I originally developed as a light-weight transactional query frontend for concept-oriented terminologies. At first, it implemented a query-language interface (TQL), but I later modified to use URL based commands, which allowed me to leverage http/https and the web server as transactional infrastructure.

Clinical Resources Center (CRC) (1997 - present)

I originally developed the CRC with Benny Poon, David Oaks, and Vicki Bencken of the Center for Health Technology (CHT).  We conceived of a configurable web based system that could provide basic dynamic 'pages' with links to internal (private) and external(public) web content.  At the time, we were hosting over 80 websites for various groups within the UC Davis Health System and wanted to provide an easy alternative to traditional hosting. Today, these systems are known as "portal systems".  We took the concept further and used the system to implement a knowledge portal to provide clinically relevant reference information at the point of care using a web interface. Our first such resource was Gold's drug database, a searchable ICD database, and a searchable formulary.  Today, it continues to be used and has about 5,000 users throughout UCDHS. It provides access to a large array of resources including online web-enabled textbooks, internally authored clinical guidelines, and residency curricula for many departments.  More... CRC Presentation: Distributing Medical Knowledge Content to the Healthcare Enterprise  [2001]

QD-POP - Quick and Dirty POP (1996 - 2000)

Designed and developed in 1996, it was one of the first systems to web-enable email. Jose Galvez, David Hill, and myself developed the system using Perl, a commonly used language in  common gateway interface (CGI) programming at the time. CGI was a way to web-enable access to backend non-web systems.  Amazingly, QD-POP continues to be used today -- nearly 15 years later! 

ODYSSEY (1995 - 1997)

A project involving the development of a specialized web browser for use in healthcare environments. This was early in the evolution of the Web and mostly an experiment in using Web-based technology in a healthcare setting. Designed by Benny Poon and myself, Odyssey was a configurable web browser built in Delphi (Object Pascal on Windows) implementing a number of features not present in Mosaic, the most used browser at the time.  We implemented 'quick links' on the browser's URL bar to enable quick access to Medline, QD-POP (see above) and selected links to high-value healthcare web content (the Virtual Hospital, etc..).  Most browsers today support this feature as a bookmark-bar. The browser's default behavior could be configured using a local administrative password. For example, one could disable the print and openURL functions.  We proposed that browsers should have this kind of configurability "on the fly" after an authentication step with a remote web server, which would provide the configuration based on the user and location using information embedded in the HTTP header using META tags.