2017 meant change for the Text Information Extraction System (TIES) software. We released two new versions (5.65 and 5.7), and TIES started processing documents for Stony Brook University. Additionally, TIES and the TCRN members published several papers on cancer research and Noble Coder (a natural language processing component). TIES also gained a new TCRN website, designed to make becoming a new TIES node even easier. All of this progress helped fuel research and continue the process of tissue exchange across the TCRN.
On February 1st, TIES added new features for data and image viewing in case sets, a new structured data view (designed to add a table that transposes structured data for attributes with multiple values), and implemented a file based query save/load system. While several bug fixes were also handled, some of the most important included the ability to change the expiration dates on a study, update user profile information, and the capacity to save/load data queries containing demographic criteria.
At the beginning of July, TIES added improved structured data loading for grouped data items and improved performance for loading large datasets. The top bug fixes for this version of the software included TCGA node duplication for Roswell Park Cancer Institute (RPCI) and the ability to copy/move multiple selected items across case sets.
Our 2017 efforts also afforded TIES the chance to expand its websites to include a TCRN page. This website gives users a rundown of the members within the TCRN, the key details about joining the network, and all of the documents associated with becoming a node. Designed for easy navigation, the website is also there to help with policy questions and include recommendations for organizations to implement the software. With the TCRN continuing to grow, the website helped keep all of the documents in one area and increase the awareness of the network for future sites to join.
Published and Accepted Papers
One of the major papers to come out of TIES this year included “A natural language processing system for extracting cancer phenotypes from clinical records.” This paper featured the precise phenotype information needed to understand the effects of genetic and epigenetic changes on tumor behavior and responsiveness. Two other important papers “Automated annotation and classification of BI-RADS assessment from radiology reports” and “The role of skin ulceration in breast carcinoma staging and outcome” were published. Several other pilot projects are ongoing, and we expect an increasing number of publications as the TCRN continues to grow.
While TIES has grown significantly in 2017, there is still a lot more to achieve in order to meet our vision of a nationwide data, image and biospecimen network that supports cancer research. Stay tuned for more exciting developments in the year 2018!