A text processing pipeline to extract recommendations from radiology reports

Communication of follow-up recommendations when abnormalities are identified on imaging studies is prone to error. The absence of an automated system to identify and track radiology recommendations is an important barrier to ensuring timely follow-up of patients especially with non-acute incidental findings on imaging examinations. In this paper, we present a text processing pipeline to automatically identify clinically important recommendation sentences in radiology reports. Our extraction pipeline is based on natural language processing (NLP) and supervised text classification methods. To develop and test the pipeline, we created a corpus of 800 radiology reports double annotated for recommendation sentences by a radiologist and an internist. We ran several experiments to measure the impact of different feature types and the data imbalance between positive and negative recommendation sentences. Our fully statistical approach achieved the best f-score 0.758 in identifying the critical recommendation sentences in radiology reports.

[1]  James H Thrall,et al.  Recommendations for additional imaging in radiology reports: multifactorial analysis of 5.9 million examinations. , 2009, Radiology.

[2]  Wenqian Shang,et al.  A novel feature selection algorithm for text categorization , 2007, Expert Syst. Appl..

[3]  Pragya A. Dang,et al.  Extraction of recommendation features in radiology with natural language processing: exploratory study. , 2008, AJR. American journal of roentgenology.

[4]  L. Lucey,et al.  The ACR guideline on communication: to be or not to be, that is the question. , 2010, Journal of the American College of Radiology : JACR.

[5]  J. Austin,et al.  Guidelines for management of small pulmonary nodules detected on CT scans: a statement from the Fleischner Society. , 2005, Radiology.

[6]  James A Brink,et al.  Managing incidental findings on abdominal CT: white paper of the ACR incidental findings committee. , 2010, Journal of the American College of Radiology : JACR.

[7]  Dan Klein,et al.  Feature-Rich Part-of-Speech Tagging with a Cyclic Dependency Network , 2003, NAACL.

[8]  Hardeep Singh,et al.  Research Paper: Communication Outcomes of Critical Imaging Results in a Computerized Notification System , 2007, J. Am. Medical Informatics Assoc..

[9]  Gordon D Schiff,et al.  Medical error: a 60-year-old man with delayed care for a renal mass. , 2011, JAMA.

[10]  Carol Friedman,et al.  Identification of findings suspicious for breast cancer based on natural language processing of mammogram reports , 1997, AMIA.

[11]  Henry J. Lowe,et al.  Selective Automated Indexing of Findings and Diagnoses in Radiology Reports , 2001, J. Biomed. Informatics.

[12]  M. L. R. D. Christenson,et al.  Guidelines for Management of Small Pulmonary Nodules Detected on CT Scans: A Statement From the Fleischner Society , 2006 .

[13]  Haibo He,et al.  Learning from Imbalanced Data , 2009, IEEE Transactions on Knowledge and Data Engineering.

[14]  James H Thrall,et al.  Application of Recently Developed Computer Algorithm for Automatic Classification of Unstructured Radiology Reports: Validation Study 1 , 2004 .

[15]  Jay A. Moskovitz,et al.  Creating a comprehensive customer service program to help convey critical and acute results of radiology studies. , 2011, AJR. American journal of roentgenology.

[16]  J Starren,et al.  Architectural requirements for a multipurpose natural language processor in the clinical environment. , 1995, Proceedings. Symposium on Computer Applications in Medical Care.

[17]  Alan R. Aronson,et al.  An overview of MetaMap: historical perspective and recent advances , 2010, J. Am. Medical Informatics Assoc..

[18]  Fei Xia,et al.  Automatic identification of critical follow-up recommendation sentences in radiology reports. , 2011, AMIA ... Annual Symposium proceedings. AMIA Symposium.

[19]  J. Austin,et al.  Use of natural language processing to translate clinical information from a database of 889,921 chest radiographic reports. , 2002, Radiology.

[20]  Carol Friedman,et al.  Research Paper: A General Natural-language Text Processor for Clinical Radiology , 1994, J. Am. Medical Informatics Assoc..

[21]  Lucy Vanderwende,et al.  Statistical Section Segmentation in Free-Text Clinical Records , 2012, LREC.

[22]  Adam L. Berger,et al.  A Maximum Entropy Approach to Natural Language Processing , 1996, CL.

[23]  James H Thrall,et al.  Addressing overutilization in medical imaging. , 2010, Radiology.

[24]  Christopher L. Roy,et al.  Patient Safety Concerns Arising from Test Results That Return after Hospital Discharge , 2005, Annals of Internal Medicine.

[25]  Leonard Berlin,et al.  Failure of radiologic communication: An increasing cause of malpractice litigation and harm to patients , 2010, Applied Radiology.

[26]  Dunja Mladenic,et al.  Feature selection on hierarchy of web documents , 2003, Decis. Support Syst..

[27]  Olivier Bodenreider,et al.  The Unified Medical Language System (UMLS): integrating biomedical terminology , 2004, Nucleic Acids Res..

[28]  Martin F. Porter,et al.  An algorithm for suffix stripping , 1997, Program.

[29]  N L Jain,et al.  Identification of suspected tuberculosis patients based on natural language processing of chest radiograph reports. , 1996, Proceedings : a conference of the American Medical Informatics Association. AMIA Fall Symposium.