The Dermatology team at NCI collects, examines and classifies images of lesions from hundreds of patients. Each patient visits the site tens of times, and at each visit up to 20 images are taken of lesions and skin conditions of the patient. Prior to the adoption of Catalyzer, the team was attempting to manage their collection of patient photos and associated information using the basic capabilities offered by Windows Explorer.

As part of a patient's consult, a professional photographer takes digital photos of the patient's lesions. The image files are placed in a folder, located at a specific shared network drive, for future classification. The research nurses or doctors would later append the diagnostic information to the folder and file names and then move the entire set to a folder with other already classified images. Locating images was accomplished by a regular Windows Explorer search on the folder/file name or by browsing the folder hierarchy. Images of interest were then copied to another folder on the drive. Storage volume eventually became a concern, as the team was storing duplicates of high resolution images.

In addition to the organization hierarchy being reliant on a manual act of relocating folders and files - a task which can cause problems in even the most disciplined of teams - the complicated set of file names that resulted often became unwieldy. There was also the age old problem of individualism - it was not unheard of for each doctor or nurse on occasion to use terminology that was slightly different than that of his/her peers. This would result in a great degree of difficulty when one individual had to track down images or data that had been generated by another. As in many other cases, often work was stalled while coworkers tried to imagine what one person may have been thinking, or waited for their colleague to return to sort out their naming conventions.

Another obstacle was the size of the images. Because of the nature of the work, the images must be of a high quality and resolution. Whenever the team was able to locate an image whose filename and folder location met specific criteria and they needed a visual confirmation, they were required to open the source image. Because of the size of these files and the applications needed to view them, this was often a slow and processor intensive function.

In short, the workflow was cumbersome. The team needed a tool that would not only address their needs to view images, but information storage, data mining, and sharing. Jeff Shilling, an IT consultant to the team, suggested they look at Catalyzer.

In the new workflow using Catalyzer, there is a single catalog per patient, a subfolder for each visit, and a record for each image. As before, captured images are uploaded to a central network drive, but that is where the similarity ends. Now instead of sorting files and adding information to the file name, the files can stay in the patient's secure network folder and remain a single, unaltered copy.

Upon completion of a patient's visit for the day, the doctors or nurses import their high resolution images into a patient catalog in Catalyzer - directly from the network drive. Catalyzer's ability to extract and display thumbnails means the Derm team does not have to store these images on their local machines, nor do they have to use complicated image suites to view them.

Once the images are imported into Catalyzer, the nurses and doctors are able to associate information about the patient, diagnosis, etc., in a database format instead of using long and complicated folder/file names.

A sampling of the information gathered with images and linked using Catlyzer includes: demographic information, diagnostic, histology, pathology, disease type, body site, lesion size, color, and margin/order characteristics. Because Catalyzer allows lead investigators to standardize templates and vocabularies for researchers to work with, there is no longer any variance in terminology. Each lesion is classified using a standard choice of terms, decided upon by the administrator during catalog design.

Practically, this standardization of terms means each user's work is subject to query and immediate use by colleagues without the need for interpretation or guesswork. There is no need for the person who recorded the information to explain their logic to others, because everyone's templates are the same and each field has help-text associated with it to aid in training of new staff. Groups of images that match specific criteria can even be exported to new catalogs, so researchers can distill what might be a large patient catalog down into a catalog consisting only of a specific subset of records they wish to focus on.

Catalyzer's robust search capabilities mean that the team can locate quickly specific images based on any of the above criteria. Because the thumbnails reside within the catalog, team members can tell immediately if they have retrieved the correct record. Using Catalyzer's image annotation tools, they can even highlight and comment upon a region of interest, saving them time when they later need to locate an exact spot that concerns them.

Catalyzer's ability to display, organize, annotate and tag images means fewer headaches for everyone. Individual doctors and nurses save valuable time, lead researchers are easily able to review and interpret the ongoing cases managed by their team, and the IT staff have little or no additional overhead work to maintain the application.