HealthLibrarian
Have you experienced a loved one suffering from a medical condition or disease that is either rare or difficult to treat? And you know that there are a myriad databases and research papers available from academic and government institutions. But how to access, and make sense of, all those sources through a single interface?
HealthLibrarian is a health-focused, patient-centered vertical search engine for health data. Our system has been architected from the ground up to present bio-medical knowledge in a consumer-friendly format. We make both Open Access and Open Data much more easy to use.
The HealthLibrarian server is functionally and operationally different from all the “horizontal search engines” currently in the market. Generic search engines (such as Google and Yahoo) present results based on text matches and on a site’s popularity. HealthLibrarian is designed to intelligently sort and categorize all the health-related information freely available from validated, scientific sources. Users can fine-tune their query to maximize the relevance of the result set.
Visit a private label version of HealthLibrarian.
For Consumers
Healthcare is Local: A user is most interested in providers and resources available in user’s immediate geographic area. HealthLibrarian presents results based on user’s location.
Healthcare is Personal: Also, users want information that is relevant to their age, sex, race, and condition or disease. HealthLibrarian runs its searches based on these demographic parameters.
For Providers
Providers unfortunately do not have enough time to research the latest information available to best treat each and every one of their patients. HealthLibrarian offers a single access point to the myriad databases with high-quality health data available from the US government. HealthLibrarian provides highly-relevant, context-specific information both users and providers can use immediately.
HealthLibrarian also offers providers the ability to run queries on all their patients through our search engine. This service could be sponsored and promoted by your organization.
About HealthLibrarian
These documents will give you a very detailed overview of HealthLibrarian's current and future capabilities:
Platform v02
We are currently developing the next generation of our software platform. Based exclusively on Free and Open Source Software ("F/OSS"), this platform is designed to crawl, parse, analyze and index massive amounts of information. Particularly Open Access materials.
By using domain-specific dictionaries we are then able to provide semantic search capabilities tailored to specific industries: health, legal, business, etc. Download Overview of HealthLibrarian's Platform v02.
Why Licensing HealthLibrarian?
Should you build your own health-data search engine or license ours? Benefits of licensing HealthLibrarian presents our point of view.
Services Available
We have 02 complementary services available right now using the same server technology. See {OpenPHI_HL_versions_05.pdf} for an overview.
HLv01
Visit HealthLibrarian's site
This is the entry point to the search engine we're licensing to our clients. This interface allows users to develop personalized "Information Prescriptions" for consumers and researchers.
HLv02
HealthLibrarian's consumer portal
This is a consumer-facing portal that leverages our search engine technology. WE currently have 10 million unique URLs about health-related issues (we will soon have over 20 million unique URLs available). We can place ads for all of your organization's health-related articles and books in the most appropriate pages of our collection.
Business Value
HealthLibrarian provides a turn-key package to companies and organizations in the healthcare space. We save our customers from maintaining their own databases of clinical trials, scientific articles, clinical guidelines, etc. We take care of all technical issues, and enhance and update the databases regularly for a flat monthly fee.
The HealthLibrarian appliance identifies, downloads, processes and indexes high-quality health-related databases from governmental and scientific sources in the US and overseas (i.e. caBIG; NIH; NLM).
Your organization can license HealthLibrarian through a few models:
- Hosted (ASP). The client runs queries from its server to HealthLibrarian's back-end server farm.
- On-premises. A HealthLibrarian appliance is installed inside the client's own network. There is absolutely no private information ever leaving the client's network.
Technical Core
This is the problem we're tackling:
- Hundreds of sources
- Multiple domain-specific ontologies
- Across agencies
- Across countries
Our solution:
- Massive, Unified Data Warehouse
- Single Controlled Medical Terminology
We have built and continue to enhance a significant Data Warehouse with multiple datasources already available.
Click image to maximize it.
All the raw data from each datasource has been downloaded and stored in OpenPHI's servers. Each record is then indexed and analyzed using our Controlled Medical Vocabulary ("CMV"), an enhanced version of the National Cancer Institute Enterprise Vocabulary Server ("EVS"). Our CMV has over 07 Million concepts. HealthLibrarian then leverages our CMV to conduct "semantic searching" to show user a list of medically-related terms. Searching for "heart" for example will also include all references to "cardiac" as well.
Click image to maximize it.
The Data Warehouse uses Free Software exclusively: GNU/Linux operating system (particularly Ubuntu 9.x); Apache web server; MySQL, a RDBMS (database engine); PHP, a web-optimized scripting language and Python, a powerful scripting language. This is known in the industry as the "LAMP stack".
The Data Warehouse provides common functionality to access, load, manage, query and retrieve information. We have also developed a re-usable architecture through open interfaces that expose the entire Data Warehouse via what is referred to as "Web Services".
The Data Warehouse is completely scalable through Amazon Web Services Elastic Compute Cloud service, is both robust and resilient, and operates at very low cost. This is known as "cloud computing".
The Data Warehouse is implemented as a series of Virtual Machine ("VM") hosted in Amazon Web Services' huge data centers. Think of a Virtual Machine ("VM") as a full-blown Internet-class server that happens to run inside a software-based "container" in another computer. For example, you can run the MS Windows operating system on an Apple Mac by running Windows as a "virtual appliance" on top of the Mac operating system. This is known as "virtualization".
HIPAA Notice: For privacy and security reasons, OpenPHI can deploy a HealthLibrarian server inside a customer's network.
HealthLibrarian v03
We're building out our vision for the next version of HealthLibrarian. See diagram below.
Click image to maximize it.
Think of this as an enhanced electronic bookshelf in the life sciences. We're in the process of loading 100s of MBs of raw data from government sources in the genetic, protein, anatomical, and community health / Public Health domains.
The end goal is to allow users to "navigate" our data warehouse through multiple axes. After selecting a disease of interest, a user will then be able to:
- drill down to the proteins associated with that disease; and then further down to the genes that express such proteins ("biological view")
- explore all the medications related to that disease; and further down to the chemical compound(s) in each such medication ("chemical view")
- see what tissue the disease is related to; and the tissue is part of an organ; and the organ is an element of a particular anatomical system ("anatomical view")
- explore the impact of community and Public Health issues in the individual diseases ("community view")
It is important to note that this massive data mining process is driven entirely by our own software tools. We are building Natural Language Processing ("NLP") and semantic search tools to extract meaning and "connections" from the raw data we have access to.
