ARTIFICIAL INTELLIGENCE: NATURAL LANGUAGE PROCESSING FOR PEER-REVIEW IN RADIOLOGY

Objective. To assess the importance of natural language processing (NLP) system for quality assurance of the radiological reports.

Material and methods. Multilateral analysis of chest low-dose computed tomography (LDCT) reports based on a commercially available cognitive NLP system was performed. The applicability of artificial intelligence for discrepancy identification in the report body and conclusion (quantitative analysis) and radiologist adherence to the Lung-RADS guidelines (qualitative analysis) was evaluated.

Results. Quantitative analysis: in the 8.3% of cases LDCT reports contained discrepancies between text body and conclusion, i.e., lung nodule described only in body or conclusion. It carries potential risks and should be taken into account when performing a radiological study audit. Qualitative analysis: for the Lung-RADS 3 nodules, the recommended principles of patient management were used in 46%, for Lung-RADS 4A – in 42%, and for Lung-RADS 4B – in 49% of cases.

Conclusion. The consistency of NLP system within the framework of radiological study audit was 95–96%. The system is applicable for the radiological study audit, i.e. large-scale automated analysis of radiological reports and other medical documents.

1. Brady A. Error and discrepancy in radiology: inevitable or avoidable? Ins. Imag.2017; 1 (8): 171–82.

2. Larson D.B., Donnelly L.F., Podberesky D.J., Merrow A.C., Sharpe R.E. Jr, Kruskal J.B. Peer feedback, learning, and improvement: answering the call of the institute of medicine report on diagnostic error 1. Radiology. 2017; 283 (1): 231–41. DOI: 10.1148/radiol.2016161254

3. Walker E.A., Petscavage-Thomas J.M., Fotos J.S., Bruno M.A. Quality metrics currently used in academic radiology departments: results of the QUALMET survey. Br. J. Radiol. 2017; 90 (1071): 20160827. DOI: 10.1259/bjr.20160827

4. Morozov S., Guseva E., Ledikhova N., Vladzymyrskyy A., Safronov D. Telemedicine-based system for quality management and peer review in radiology. Ins. Imag. 2018; 9 (3): 337–41. DOI: 10.1007/s13244-018-0629-y

5. Morozov S.P., Pereverzev M.O. Radiology – avanguard of healthcare informatization. Rossiyskiy Elektronnyy Zhurnal Luchevoy Diagnostiki (Russian Electronic Journal of Radiology).2013; 3: 41–50 (in Russ.).

6. Polishchuk N.S., Vetsheva N.N., Kosarin S.P., Morozov S.P., Kuz’mina E.S. Unified radiological information service as a key element of organizational and methodical work of Research and Practical Center of Medical Radiology. Radiologiya–Praktika (Radiology–Practice). 2018; 1 (67): 6–17 (in Russ.).

7. Gusev A.V. Prospects for neural networks and deep machine learning in creating health solutions. Vrach i Informatsionnye Tekhnologii (Information Technologies for the Physician).2017; 3: 92–105 (in Russ.)

8. Clark T.J., Flood T.F., Maximin S.T., Sachs P.B. Lung CT screening reporting and data system speed and accuracy are increased with the use of a semiautomated computer application. J. Am. Coll. Radiol. 2015; 12 (12 Pt A): 1301–6.

9. Barchuk A.A., Arsen'ev A.I., Belyaev A.M., Gombolevskiy V.A., Nefedova A.V., Kanaev S.V. et al. The effectiveness of screening for cancer. Voprosy Onkologii (Problems in Oncology). 2017; 4: 557–67 (in Russ.).

10. Gombolevskiy V.A., Barchuk A.A., Laypan A.S., Vetsheva N.N., Vladzimirskiy A.V., Morozov S.P. Lung сancer screening with low-dose computed tomography: management and efficiency. Radiologiya–Praktika (Radiology–Practice). 2018; 1 (67): 28–33 (in Russ.)

11. Gombolevskiy V.A., Kharlamov K.A., Pyatnitskiy I.A., Kim S.Yu., Morozov S.P. Patterns of describing CT protocols examination. Guidelines. 2016; 23: 13–4 (in Russ.). Available at: http://medradiology.moscow/d/1364488/d/metodicheskiye_rekomendatsii_no_23_2016_g_shablony_protokolov_opisaniy_kt_issledovaniy.pdf (accessed 20October 2018).

12. Basmanov S.N., Basmanova A.A. Areview of the evolution of expert systems in medicine in terms of compliance with the main features. Perspektivy Razvitiya Informatsionnykh Tekhnologiy. 2014; 21: 126–30 (in Russ.).

13. Bulu H., Sippo D.A., Lee J.M., Burnside E.S., Rubin D.L. Proposing new radlex terms by analyzing free-text mammography reports. J. Digit. Imag. 2018. DOI: 10.1007/s10278-018-0064-0

14. Goff D.J., Loehfelm T.W. Automated radiology report summarization using an open-source natural language processing pipeline. J. Digit. Imag.2018; 31 (2): 185–92. DOI: 10.1007/s10278-017-0030-2

15. Huesch M.D., Cherian R., Labib S., Mahraj R. Evaluating report text variation and informativeness: natural language processing of CT chest imaging for pulmonary embolism. J. Am. Coll. Radiol. 2018; 15 (3 Pt B): 554–62. DOI: 10.1016/j.jacr.2017.12.017

16. Kahn C.E. Jr. An ontology-based approach to estimate the frequency of rare diseases in narrative-text radiology reports. Stud. Health. Technol. Inform. 2017; 245: 896–900.

17. Percha B., Zhang Y., Bozkurt S., Rubin D., Altman R.B., Langlotz C.P. Expanding a radiology lexicon using contextual patterns in radiology reports. J. Am. Med. Inform. Assoc.2018; 10. DOI: 10.1093/jamia/ocx152

18. Pons E., Braun LM., Hunink M.G., Kors J.A. Natural language processing in radiology: a systematic review. Radiology. 2016; 279 (2): 329–43. DOI: 10.1148/radiol.16142770

19. Gálvez J.A., Pappas J.M., Ahumada L., Martin J.N., Simpao A.F., Rehman M.A. et al. The use of natural language processing on pediatric diagnostic radiology reports in the electronic health record to identify deep venous thrombosis in children. J. Thromb. Thrombolysis. 2017. DOI: 10.1007/s11239-017-1532-y

20. Tian Z., Sun S., Eguale T., Rochefort C.M. Automated extraction of VTE events from narrative radiology reports in electronic health records: a validation study. Med. Care. 2017; 55 (10): e73–e80. DOI: 10.1097/MLR.0000000000000346

21. Trivedi H., Mesterhazy J., Laguna B., Vu T., Sohn J.H. Automatic determination of the need for intravenous contrast in musculoskeletal MRI examinations using IBM Watson's natural language processing algorithm. J. Digit. Imag. 2017; 18. DOI: 10.1007/s10278-017-0021-3

22. Guimaraes C.V., Grzeszczuk R., Bisset G.S. 3rd, Donnelly L.F. Comparison between manual auditing and a natural language process with machine learning algorithm to evaluate faculty use of standardized reports in radiology. J. Am. Coll. Radiol.2018; 15 (3 Pt B): 550–3. DOI: 10.1016/j.jacr.2017.10.042

23. Humphrey L.L., Deffebach M., Pappas M., Zakher B., Fu R., Slatore C.G. et al. Screening for lung cancer with lowdose computed tomography: a systematic review to update the U.S. Preventive services task force recommendation. Ann. Int. Med. 2013; 159 : 411–20.

24. Oudkerk M., Devaraj A., Vliegenthart R., Henzler T., Prosch H., Heussel C.P. European position statement on lung cancer screening. Lancet Oncol. 2017; 18 (12): 754–66.

25. Beyer S.E., McKee B.J., Regis S.M., McKee A.B., Flacke S., El Saadawi G. et al. Automatic LungRADS™ classification with a natural language processing system. J. Thorac. Dis. 2017; 9 (9): 3114–22. DOI: 10.21037/jtd.2017.08.13

26. Lacson R., Prevedello L.M., Andriole K.P., Gill R., LenociEdwards J., Roy C. et al. Factors associated with radiologists' adherence to Fleischner Society guidelines for management of pulmonary nodules.J. Am. Coll. Radiol. 2012; 9 (7): 468–73. DOI: 10.1016/j.jacr.2012.03.009