Rationale and Objectives
Adoption of the Prostate Imaging Reporting & Data System (PI-RADS) has been shown
to increase detection of clinically significant prostate cancer on prostate mpMRI.
We propose that a rule-based algorithm based on Regular Expression (RegEx) matching
can be used to automatically categorize prostate mpMRI reports into categories as
a means by which to assess for opportunities for quality improvement.
Materials and Methods
All prostate mpMRIs performed in the Duke University Health System from January 2,
2015, to January 29, 2021, were analyzed. Exclusion criteria were applied, for a total
of 5343 male patients and 6264 prostate mpMRI reports. These reports were then analyzed
by our RegEx algorithm to be categorized as PI-RADS 1 through PI-RADS 5, Recurrent
Disease, or “No Information Available.” A stratified, random sample of 502 mpMRI reports
was reviewed by a blinded clinical team to assess performance of the RegEx algorithm.
Results
Compared to manual review, the RegEx algorithm achieved overall accuracy of 92.6%,
average precision of 88.8%, average recall of 85.6%, and F1 score of 0.871. The clinical
team also reviewed 344 cases that were classified as “No Information Available,” and
found that in 150 instances, no numerical PI-RADS score for any lesion was included
in the impression section of the mpMRI report.
Conclusion
Rule-based processing is an accurate method for the large-scale, automated extraction
of PI-RADS scores from the text of radiology reports. These natural language processing
approaches can be used for future initiatives in quality improvement in prostate mpMRI
reporting with PI-RADS.
Keywords
Abbreviations:
mpMRI (multiparametric magnetic resonance imaging), csPCa (clinically significant prostate cancer), PI-RADS (Prostate Imaging Reporting & Data System), NLP (natural language processing), AI (artificial intelligence), ML (machine learning), RegEx (regular expression), QI (quality improvement)To read this article in full you will need to make a payment
Purchase one-time access:
Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online accessOne-time access price info
- For academic or personal research use, select 'Academic and Personal'
- For corporate R&D use, select 'Corporate R&D Professionals'
Subscribe:
Subscribe to Academic RadiologyAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
References
Siegel RL, Miller KD, Fuchs HE, et al.Cancer statistics, 2021. CA Cancer J Clin2021;71(1):7-33. doi:10.3322/caac.21654
- Dramatic increase in the utilization of multiparametric magnetic resonance imaging for detection and management of prostate cancer.Abdom Radiol (NY). 2017; 42: 1255-1258https://doi.org/10.1007/s00261-016-0975-5
- Prostate imaging reporting and data system version 2.1: 2019 update of prostate imaging reporting and data system version 2.Eur Urol. 2019; 76: 340-351https://doi.org/10.1016/j.eururo.2019.02.033
- Diagnostic performance of prostate imaging reporting and data system version 2 for detection of prostate cancer: a systematic review and diagnostic meta-analysis.Eur Urol. 2017; 72: 177-188https://doi.org/10.1016/j.eururo.2017.01.042
- Accuracy and agreement of PIRADSv2 for prostate cancer mpMRI: a multireader study.J Magn Reson Imaging. 2017; 45: 579-585https://doi.org/10.1002/jmri.25372
- Interreader variability of prostate imaging reporting and data system version 2 in detecting and assessing prostate cancer lesions at prostate MRI.AJR Am J Roentgenol. 2019; : 1-8https://doi.org/10.2214/ajr.18.20536
- Intra- and interreader reproducibility of PI-RADSv2: a multireader study.J Magn Reson Imaging. 2019; 49: 1694-1703https://doi.org/10.1002/jmri.26555
- Prostate magnetic resonance imaging interpretation varies substantially across radiologists.Eur Urol Focus. 2019; 5: 592-599https://doi.org/10.1016/j.euf.2017.11.010
- Factors influencing variability in the performance of multiparametric magnetic resonance imaging in detecting clinically significant prostate cancer: a systematic literature review.Eur Urol Oncol. 2020; 3: 145-167https://doi.org/10.1016/j.euo.2020.02.005
- Application of rostate maging eporting and ata ystem version 2 (PI-RADS v2): interobserver agreement and positive predictive value for localization of intermediate- and high-grade prostate cancers on multiparametric magnetic resonance imaging.Acad Radiol. 2017; 24: 1101-1106https://doi.org/10.1016/j.acra.2017.03.019
- Interobserver agreement and positivity of PI-RADS version 2 among radiologists with different levels of experience.Acad Radiol. 2019; 26: 1017-1022https://doi.org/10.1016/j.acra.2018.08.013
- Improving communication of diagnostic radiology findings through structured reporting.Radiology. 2011; 260: 174-181https://doi.org/10.1148/radiol.11101913
- From guidelines to practice: how reporting templates promote the use of radiology practice guidelines.J Am Coll Radiol. 2013; 10: 268-273https://doi.org/10.1016/j.jacr.2012.09.025
- Impact of a structured reporting template on adherence to prostate imaging reporting and data system version 2 and on the diagnostic performance of prostate MRI for clinically significant prostate cancer.J Am Coll Radiol. 2018; 15: 749-754https://doi.org/10.1016/j.jacr.2018.01.034
- A comparison of radiologists' and urologists' opinions regarding prostate MRI reporting: results from a survey of specialty societies.AJR Am J Roentgenol. 2018; 210: 101-107https://doi.org/10.2214/ajr.17.18241
- Innovative standardized reporting template for prostate mpMRI improves clarity and confidence in the report.World J Urol. 2021; 39: 2447-2452https://doi.org/10.1007/s00345-020-03487-3
- Prostate MRI qualification: AJR expert panel narrative review.AJR Am J Roentgenol. 11 2022; https://doi.org/10.2214/ajr.22.27615
- Extracting information from textual documents in the electronic health record: a review of recent research.Yearb Med Inform. 2008; : 128-144
- Practical guide to natural language processing for radiology.Radiographics. 2021; 41: 1446-1453https://doi.org/10.1148/rg.2021200113
- Automated extraction of BI-RADS final assessment categories from radiology reports with natural language processing.J Digit Imaging. 2013; 26: 989-994https://doi.org/10.1007/s10278-013-9616-5
- Natural language processing to identify pulmonary nodules and extract nodule characteristics from radiology reports.Chest. 2021; 160: 1902-1914https://doi.org/10.1016/j.chest.2021.05.048
- Basic artificial intelligence techniques: natural language processing of radiology reports.Radiol Clin North Am. 2021; 59: 919-931https://doi.org/10.1016/j.rcl.2021.06.003
- Learning regular expressions for clinical text classification.J Am Med Inform Assoc. 2014; 21: 850-857https://doi.org/10.1136/amiajnl-2013-002411
- Prostate imaging and data reporting system version 2 as a radiology performance metric: an analysis of 18 abdominal radiologists.J Am Coll Radiol. 2021; 18: 1069-1076https://doi.org/10.1016/j.jacr.2021.02.032
- PI-RADS prostate imaging - reporting and data system: 2015, version 2.Eur Urol. 2016; 69: 16-40https://doi.org/10.1016/j.eururo.2015.08.052
R: A language and environment for statistical computing.R foundation for statistical computing; 2021. Available at: https://www.R-project.org/. Accessed 20 May 2022.
- Imaging facilities' adherence to PI-RADS v2 minimum technical standards for the performance of prostate MRI.Acad Radiol. 2018; 25: 188-195https://doi.org/10.1016/j.acra.2017.08.013
- Quality of prostate MRI: is the PI-RADS standard sufficient?.Acad Radiol. 2021; 28: 199-207https://doi.org/10.1016/j.acra.2020.01.031
- A critical analysis of deficiencies in the quality of information contained in prostate multiparametric MRI requests and reports.Ir J Med Sci. 2022; https://doi.org/10.1007/s11845-021-02875-x
- Use of a quality improvement initiative to achieve consistent reporting of level of suspicion for tumor on multiparametric prostate MRI.AJR Am J Roentgenol. 2016; 206: 1040-1044https://doi.org/10.2214/ajr.15.15768
- The Trials and Tribulations of Assembling Large Medical Imaging Datasets for Machine Learning Applications.J Digit Imaging. 2021; 34: 1424-1429https://doi.org/10.1007/s10278-021-00505-7
- Classification of multiple diseases on body CT scans using weakly supervised deep learning.Radiol Artif Intell. 2022; 4e210026https://doi.org/10.1148/ryai.210026
- Multi-label annotation of text reports from computed tomography of the chest, abdomen, and pelvis using deep learning.BMC Med Inform Decis Mak. 2022; 22: 102https://doi.org/10.1186/s12911-022-01843-4
- Development and implementation of a multisite registry using structured templates for actionable findings in the kidney.J Am Coll Radiol. 2022; https://doi.org/10.1016/j.jacr.2022.02.030
Article info
Publication history
Published online: July 28, 2022
Accepted:
June 28,
2022
Received in revised form:
June 15,
2022
Received:
May 21,
2022
Footnotes
Competing Interests: Rajan T. Gupta, MD – Consultant, Invivo Corp.
No other financial disclosures for all other authors.
Identification
Copyright
© 2022 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.
