Artificial neural network modeling enhances risk stratification and can reduce downstream testing for patients with suspected acute coronary syndromes, negative cardiac biomarkers, and normal ECGs
- 354 Downloads
Despite uncertain yield, guidelines endorse routine stress myocardial perfusion imaging (MPI) for patients with suspected acute coronary syndromes, unremarkable serial electrocardiograms, and negative troponin measurements. In these patients, outcome prediction and risk stratification models could spare unnecessary testing. This study therefore investigated the use of artificial neural networks (ANN) to improve risk stratification and prediction of MPI and angiographic results. We retrospectively identified 5354 consecutive patients referred from the emergency department for rest-stress MPI after serial negative troponins and normal ECGs. Patients were risk stratified according to thrombolysis in myocardial infarction (TIMI) scores, ischemia was defined as >5 % reversible perfusion defect, and obstructive coronary artery disease was defined as >50 % angiographic obstruction. For ANN, the network architecture employed a systematic method where the number of neurons is changed incrementally, and bootstrapping was performed to evaluate the accuracy of the models. Compared to TIMI scores, ANN models provided improved discriminatory power. With regards to MPI, an ANN model could reduce testing by 59 % and maintain a 96 % negative predictive value (NPV) for ruling out ischemia. Application of an ANN model could also avoid 73 % of invasive coronary angiograms while maintaining a 98 % NPV for detecting obstructive CAD. An online calculator for clinical use was created using these models. The ANN models improved risk stratification when compared to the TIMI score. Our calculator could also reduce downstream testing while maintaining an excellent NPV, though further study is needed before the calculator can be used clinically.
KeywordsArtificial neural networks Thrombolysis in myocardial infarction score Single-photon emission computed tomography Myocardial perfusion imaging
Compliance with ethical standards
Conflict of interest
Author Isma’eel HA declares that he has no conflict of interest. Author Cremer PC declares that he has no conflict of interest. Author Khalaf S declares that she has no conflict of interest. Author Almedawar MM declares that he has no conflict of interest. Author Elhajj IH declares that he has no conflict of interest. Author Sakr GE declares that he has no conflict of interest. Author Jaber WA declares that he has no conflict of interest.
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent was obtained from all individual participants included in the study.
- 1.CDC (2011) National Hospital Ambulatory Medical Care Survey: 2011 Emergency Department Summary Tables. United States Centers for Disease Control and Prevention National Health Care SurveysGoogle Scholar
- 2.Amsterdam EA, Kirk JD, Bluemke DA, Diercks D, Farkouh ME, Garvey JL, Kontos MC, McCord J, Miller TD, Morise A, Newby LK, Ruberg FL, Scordo KA, Thompson PD, American Heart Association Exercise CR, Prevention Committee of the Council on Clinical Cardiology CoCN, Interdisciplinary Council on Quality of C, Outcomes R (2010) Testing of low-risk patients presenting to the emergency department with chest pain: a scientific statement from the American Heart Association. Circulation 122(17):1756–1776. doi: 10.1161/CIR.0b013e3181ec61df CrossRefPubMedPubMedCentralGoogle Scholar
- 8.Hermann LK, Newman DH, Pleasant WA, Rojanasarntikul D, Lakoff D, Goldberg SA, Duvall WL, Henzlova MJ (2013) Yield of routine provocative cardiac testing among patients in an emergency department-based chest pain unit. JAMA Intern Med 173(12):1128–1133. doi: 10.1001/jamainternmed.2013.850 CrossRefPubMedGoogle Scholar
- 9.Cremer PC, Khalaf S, Agarwal S, Mayer-Sabik E, Ellis SG, Menon V, Cerqueira MD, Jaber WA (2014) Myocardial perfusion imaging in emergency department patients with negative cardiac biomarkers: yield for detecting ischemia, short-term events, and impact of downstream revascularization on mortality. Circ Cardiovasc Imaging 7(6):912–919. doi: 10.1161/CIRCIMAGING.114.002401 CrossRefPubMedGoogle Scholar
- 10.Foy AJ, Liu G, Davidson WR, Sciamanna C, Leslie DL (2015) Comparative effectiveness of diagnostic testing strategies in emergency department patients with chest pain: an analysis of downstream testing, interventions, and outcomes. JAMA Intern Med 175(3):428–436. doi: 10.1001/jamainternmed.2014.7657 CrossRefPubMedPubMedCentralGoogle Scholar
- 12.Pollack CV Jr, Sites FD, Shofer FS, Sease KL, Hollander JE (2006) Application of the TIMI risk score for unstable angina and non-ST elevation acute coronary syndrome to an unselected emergency department chest pain population. Acad Emerg Med 13(1):13–18. doi: 10.1197/j.aem.2005.06.031 CrossRefPubMedGoogle Scholar
- 13.Freeman RV, Eagle KA, Bates ER, Werns SW, Kline-Rogers E, Karavite D, Moscucci M (2000) Comparison of artificial neural networks with logistic regression in prediction of in-hospital death after percutaneous transluminal coronary angioplasty. Am Heart J 140(3):511–520. doi: 10.1067/mhj.2000.109223 CrossRefPubMedGoogle Scholar
- 15.Hagan MT, Demuth HB, Beale MH (1996) Neural network design, 1st edn. PWS Publishing Co, BostonGoogle Scholar
- 24.Pitts SR, Niska RW, Xu J, Burt CW (2008) National hospital ambulatory medical care survey: 2006 emergency department summary. Natl Health Stat Rep 7:1–38Google Scholar
- 26.deFilippi CR, Rosanio S, Tocchi M, Parmar RJ, Potter MA, Uretsky BF, Runge MS (2001) Randomized comparison of a strategy of predischarge coronary angiography versus exercise testing in low-risk patients in a chest pain unit: in-hospital and long-term outcomes. J Am Coll Cardiol 37(8):2042–2049CrossRefPubMedGoogle Scholar
- 29.Udelson JE, Beshansky JR, Ballin DS, Feldman JA, Griffith JL, Handler J, Heller GV, Hendel RC, Pope JH, Ruthazer R, Spiegler EJ, Woolard RH, Selker HP (2002) Myocardial perfusion imaging for evaluation and triage of patients with suspected acute cardiac ischemia: a randomized controlled trial. JAMA 288(21):2693–2700CrossRefPubMedGoogle Scholar
- 31.Farkouh ME, Smars PA, Reeder GS, Zinsmeister AR, Evans RW, Meloy TD, Kopecky SL, Allen M, Allison TG, Gibbons RJ, Gabriel SE (1998) A clinical trial of a chest-pain observation unit for patients with unstable angina. Chest pain evaluation in the emergency room (CHEER) investigators. N Engl J Med 339(26):1882–1888. doi: 10.1056/NEJM199812243392603 CrossRefPubMedGoogle Scholar