STRUCTURING RADIOLOGY AT SCALE: MACHINE-LEARNING CLASSIFICATION OF CTA/CTV REPORTS FOR PE/DVT AND INCIDENTAL FINDINGS

Main Article Content

Dr Lakshmi Sindhura Adusumilli
N Prajeeth Rao
Nagarjuna Gaddam

Keywords

CTA; CTV; pulmonary embolism; deep vein thrombosis; natural language processing; incidental findings

Abstract

The core of assessing the suspected cases of pulmonary embolism (PE) and deep vein thrombosis (DVT) is CT pulmonary angiography (CTA) and CT venography (CTV), but free-text reporting does not allow population-wide analytics and consistent monitoring of its incidental findings. The natural language processing (NLP) and machine-learning pipeline that we designed and tested was used to structure the CTA/CTV radiology reports, thromboembolism outcome classification, and unveil clinically interesting incidentalomas. The annotation of a de-identified corpus was done with the help of a schema that included entities (anatomy, thromboembolic patterns), relations (Location_of), and modalities (positive/negative/known/incidental/hypothetical). Plain text, concept/relationship annotations, and section typing feature sets were all trained on the Naive Bayes and Maximum Entropy models. Precision, recall and F-measure were used to measure performance. CTA+/CTV+/CTV− 24.8% vs CTA+/CTV−/CTV− 18.2% complementary diagnostic yield was emphasized in 5,000 reports CTA−/CTV+ 10.4%. Incidental findings that are of clinical significance were found in 32.0 percent of examinations. Entity annotation was very agreeable but relation extraction was relatively difficult. Baseline plain-text modeling was showing PE with F-measure 0.78 but worse in DVT and incidentalomas. Compounding concepts, modalities and relations led to significant accuracy gains: Naive Bayes made gains across categories and maximum entropy made gains to a maximum of 0.98 on thromboembolic classification and 0.80 on incidentalomas with section typing and critical-section features. Transforming narrative CTA/CTV reports into formalized representations allows to classify PE/DVT accurately and reliably, as well as identify important incidental findings. The framework advocates a semiautomated cohort introduction, choice assistance, and track-down, and elucidates the discerning added worth of CTV in high-risk groups (e.g., ICU, postoperative, malignancy, postpartum). Future work priorities are better relation/event modeling, expanding multilingual imaging lexicon and future assessing the impact of the intervention on time-to-action, following up recommendations adherence and patient outcomes. As demonstrated by this work, scalable narrative radiology based on integrating CTA/CTV data with NLP and machine learning can be converted into actionable clinical intelligence. 

Abstract 28 | Pdf Downloads 7

References

1. Loud PA, Grossman ZD, Klippenstein DL, Ray CE. (1998). Combined CT venography and pulmonary angiography: a new diagnostic technique for suspected thromboembolic disease. AJR Am J Roentgenol. 170, 951–954.
2. Goodman LR, Sostman HD, Stein PD, Woodard PK. (2009). CT venography: a necessary adjunct to ct pulmonary angiography or a waste of time, money, and radiation? 1. Radiology 250, 327–330.
3. Stein PD, Fowler SE, Goodman LR, Gottschalk A, Hales CA, Hull RD, Leeper KV, Jr, Popovich J, Jr, Quinn DA, Sos TA, Sostman HD, Tapson VF, Wakefield TW, Weg JG, Woodard PK. (2006). Multidetector computed tomography for acute pulmonary embolism. N Engl J Med. 354, 2317–2327.
4. Ghaye B, Nchimi A, Noukoua CT, Dondelinger RF. (2006). Does multi–detector row CT pulmonary angiography reduce the incremental value of indirect CT venography compared with single–detector row CT pulmonary angiography?1. Radiology. 240, 256–262.
5. Krishan S, Panditaratne N, Pandiratne N, Verma R, Robertson R. (2011). Incremental value of CT venography combined with pulmonary CT angiography for the detection of thromboembolic disease: systematic review and meta-analysis. AJR Am J Roentgenol. 196, 1065–1072.
6. Perrier A, Bounameaux H. (2006). Accuracy or outcome in suspected pulmonary embolism. N Engl J Med. 354, 2383–2385.
7. Hunsaker AR, Zou KH, Poh AC, Trotman-Dickenson B, Jacobson FL, Gill RR, Goldhaber SZ. (2008). Routine pelvic and lower extremity CT venography in patients undergoing pulmonary CT angiography. AJR Am J Roentgenol. 190, 322–326.
8. Revel MP, Sanchez O, Dechoux S, Couchon S, Frija G, Cazejust J, Chatellier G, Meyer G. (2008). Contribution of indirect computed tomographic venography to the diagnosis of postpartum venous thromboembolism. J ThrombHaemost. 6, 1478–1481.
9. Revel MP, Petrover D, Hernigou A, Lefort C, Meyer G, Frija G. (2005) . Diagnosing pulmonary embolism with four-detector row helical CT: prospective evaluation of 216 outpatients and inpatients. Radiology. 234, 265–273.
10. Hall WB, Truitt SG, Scheunemann LP, Shah SA, Rivera MP, Parker LA, Carson SS. (2009) .The prevalence of clinically relevant incidental findings on chest computed tomographic angiograms ordered to diagnose pulmonary embolism. Arch Intern Med, 169, 1961–1965.
11. Jacobs PCA, Mali WPTM, Grobbee DE, van der Graaf Y. (2008), Prevalence of incidental findings in computed tomographic screening of the chest: a systematic review. J Comput Assist Tomogr, 32, 214–221.
12. Berland LL, Silverman SG, Gore RM, Mayo-Smith WW, Megibow AJ, Yee J, Brink JA, Baker ME, Federle MP, Foley WD, Francis IR, Herts BR, Israel GM, Krinsky G, Platt JF, Shuman WP, Taylor AJ. (2010). Managing incidental findings on abdominal CT: white paper of the ACR incidental findings committee. J Am CollRadio ,7, 754–773.
13. Demner-Fushman D, Chapman WW, McDonald CJ. (2009).What can natural language processing do for clinical decision support? J Biomed Inform, 42, 760–772.
14. Friedman C, Alderson PO, Austin JH, Cimino JJ, Johnson SB. (1994). A general natural-language text processor for clinical radiology. J Am Med Inform Assoc, 1, 161–174.
15. Friedman C, Johnson SB, Forman B, Starren J. (1995). Proc AnnuSympComputAppl Med Care. Architectural requirements for a multipurpose natural language processor in the clinical environment; pp. 347–351
16. Savova GK, Masanz JJ, Ogren PV, Zheng J, Sohn S, Kipper-Schuler KC, Chute CG. (2010). Mayo clinical Text Analysis and Knowledge Extraction System (cTAKES): architecture, component evaluation and applications. J Am Med Inform Assoc, 17, 507–513.