Navigation Toggle

Case Study: Thunderstone Helps JMI Manage its Large Collection of Medical Documents

April 3, 2018
Case Study: Thunderstone Helps JMI Manage its Large Collection of Medical Documents

Who We Helped

The Jeghers Medical Index (JMI) is a medical library in St. Elizabeth Youngstown Hospital (SEYH). The JMI maintains Dr. Harold Jeghers’ collection of approximately one million historical medical journal articles that date from the late 1800s to the 1990s.

Their Challenge

After decades of amassing medical articles, JMI’s collection had grown to roughly five million physical pages, which filled more than 165 cabinets. As JMI began to digitize its collection, it was apparent that a flexible data management system that specialized for storage and searching was necessary. The professionals at JMI considered and compared several data management vendors on the following criteria:

  • Cost
  • Relevance
  • User interface
  • Methodology
  • Location of corporate headquarters
  • Willingness to customize software

After weighing all these factors, they decided that Thunderstone and TEXIS, its fully-integrated, full text search engine software platform, was the best fit for their project.

Thunderstone’s Solution

The TEXIS system is capable of hosting and managing a large collection of documents online. For JMI, TEXIS manages roughly one million articles.

With this much available information, Thunderstone wanted to provide several ways for users to find what they needed from JMI. This process included a custom search to match OCR information from hand annotated copies of articles to add key elements like PubMed XML citations and a table of anatomic terms that Jeghers created for TEXIS so that their users could find relevant articles from the original collection. Thunderstone also programmed TEXIS to offer JMI users three different search options to assist in article retrieval.

Boolean Logic

This option is patterned around search and retrieval behaviors like PubMed, an archive of biomedical and life sciences journal literature at the U.S. National Institutes of Health's National Library of Medicine. Users can use this search method to query based on the author, title, journal, abstract, and medical subject headings (MeSH).

Proximity Search

The proximity search option retrieves articles based on the occurrences of terms and location of words in sentences, paragraphs, pages, or documents. Proximity search can extract concepts contained in articles by using the relationships between words and the clustering of words with concepts to its advantage.

Jeghers Indexing

A third search option is based on the indexing method developed by Dr. Jeghers himself. This search relies on concepts such as disease, organ system, physiology, and anatomy to retrieve folders containing articles close to the search subjects.

After users made a query, the display results offered more options to organize the bibliographic data if they wanted to pare down their search even further. At this point, users can list results by various facets, including year of publication, study type, or with or without abstracts. Once a document or documents are found, the user may request a fair use copy from a JMI librarian.

JMI, a medical index powered by Thunderstone search engine software.

What Happened

As a collection of historical medical articles, the JMI allows medical professionals, professors, and students to glean more information about the past. Since the addition of TEXIS, JMI has noticed an increase in usage of the collection, with more than 2,000 hits or searches recorded over the course of a year. JMI searches are not restricted to SEYH staff, as the database is utilized by people across the world.

While the project took hundreds of hours of work to gather and upload the collection, TEXIS now has helped JMI turn into a compact, low-maintenance collection that is curated by a single full-time librarian. Since shifts in health care terms, meanings, and concepts can impact retrieval accuracy, Thunderstone can make ongoing changes to JMI’s TEXIS platform to help it improve results retrieval over time.

Work with Thunderstone Today

If your organization has a big project that requires accurate, effective search results, Thunderstone can help. Learn more about our search engine software and appliances and contact us today to talk to one of our experts about how Thunderstone can help you solve your search problems.