Attila Feher1,2, Konrad Pieszko3, Robert Miller3,4, Mark Lemley3, Aakash Shanbhag3, Cathleen Huang3, Leonidas Miras5, Yi-Hwa Liu6, Albert J Sinusas6,7, Edward J Miller6, Piotr J Slomka3. 1. Section of Cardiovascular Medicine, Department of Internal Medicine, Yale School of Medicine, Dana 3, P.O. Box 208017, New Haven, CT, 06520, USA. attila.feher@yale.edu. 2. Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT, USA. attila.feher@yale.edu. 3. Departments of Medicine (Division of Artificial Intelligence in Medicine), Biomedical Sciences and Imaging, Cedars-Sinai Medical Center, Los Angeles, CA, USA. 4. Department of Cardiac Sciences, University of Calgary, Calgary, AB, Canada. 5. Division of Cardiology, Bridgeport Hospital, Yale University School of Medicine, Bridgeport, CT, USA. 6. Section of Cardiovascular Medicine, Department of Internal Medicine, Yale School of Medicine, Dana 3, P.O. Box 208017, New Haven, CT, 06520, USA. 7. Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT, USA.
Abstract
BACKGROUND: Machine learning (ML) has been previously applied for prognostication in patients undergoing SPECT myocardial perfusion imaging (MPI). We evaluated whether including attenuation CT coronary artery calcification (CAC) scoring improves ML prediction of major adverse cardiovascular events (MACE) in patients undergoing SPECT/CT MPI. METHODS: From the REFINE SPECT Registry 4770 patients with SPECT/CT performed at a single center were included (age: 64 ± 12 years, 45% female). ML algorithm (XGBoost) inputs were clinical risk factors, stress variables, SPECT imaging parameters, and expert-observer CAC scoring using CT attenuation correction scans performed to obtain CT attenuation maps. The ML model was trained and validated using tenfold hold-out validation. Receiver Operator Characteristics (ROC) curves were analyzed for prediction of MACE. MACE-free survival was evaluated with standard survival analyses. RESULTS: During a median follow-up of 24.1 months, 475 patients (10%) experienced MACE. Higher area under the ROC curve for MACE was observed with ML when CAC scoring was included (CAC-ML score, 0.77, 95% confidence interval [CI] 0.75-0.79) compared to ML without CAC (ML score, 0.75, 95% CI 0.73-0.77, P = .005) and when compared to CAC score alone (0.71, 95% CI 0.68-0.73, P < .001). Among clinical, imaging, and stress parameters, CAC score had highest variable importance for ML. On survival analysis patients with high CAC-ML score (> 0.091) had higher event rate when compared to patients with low CAC-ML score (hazard ratio 5.3, 95% CI 4.3-6.5, P < .001). CONCLUSION: Integration of attenuation CT CAC scoring improves the predictive value of ML risk score for MACE prediction in patients undergoing SPECT MPI.
BACKGROUND: Machine learning (ML) has been previously applied for prognostication in patients undergoing SPECT myocardial perfusion imaging (MPI). We evaluated whether including attenuation CT coronary artery calcification (CAC) scoring improves ML prediction of major adverse cardiovascular events (MACE) in patients undergoing SPECT/CT MPI. METHODS: From the REFINE SPECT Registry 4770 patients with SPECT/CT performed at a single center were included (age: 64 ± 12 years, 45% female). ML algorithm (XGBoost) inputs were clinical risk factors, stress variables, SPECT imaging parameters, and expert-observer CAC scoring using CT attenuation correction scans performed to obtain CT attenuation maps. The ML model was trained and validated using tenfold hold-out validation. Receiver Operator Characteristics (ROC) curves were analyzed for prediction of MACE. MACE-free survival was evaluated with standard survival analyses. RESULTS: During a median follow-up of 24.1 months, 475 patients (10%) experienced MACE. Higher area under the ROC curve for MACE was observed with ML when CAC scoring was included (CAC-ML score, 0.77, 95% confidence interval [CI] 0.75-0.79) compared to ML without CAC (ML score, 0.75, 95% CI 0.73-0.77, P = .005) and when compared to CAC score alone (0.71, 95% CI 0.68-0.73, P < .001). Among clinical, imaging, and stress parameters, CAC score had highest variable importance for ML. On survival analysis patients with high CAC-ML score (> 0.091) had higher event rate when compared to patients with low CAC-ML score (hazard ratio 5.3, 95% CI 4.3-6.5, P < .001). CONCLUSION: Integration of attenuation CT CAC scoring improves the predictive value of ML risk score for MACE prediction in patients undergoing SPECT MPI.
Authors: Cvetan Trpkov; Alexei Savtchenko; Zhiying Liang; Patrick Feng; Danielle A Southern; Stephen B Wilton; Matthew T James; Erin Feil; Ilias Mylonas; Robert J H Miller Journal: Int J Cardiol Heart Vasc Date: 2021-06-19
Authors: Evann Eisenberg; Robert J H Miller; Lien-Hsin Hu; Richard Rios; Julian Betancur; Peyman Azadani; Donghee Han; Tali Sharir; Andrew J Einstein; Sabahat Bokhari; Mathews B Fish; Terrence D Ruddy; Philipp A Kaufmann; Albert J Sinusas; Edward J Miller; Timothy M Bateman; Sharmila Dorbala; Marcelo Di Carli; Joanna X Liang; Yuka Otaki; Balaji K Tamarappoo; Damini Dey; Daniel S Berman; Piotr J Slomka Journal: J Nucl Cardiol Date: 2021-07-06 Impact factor: 3.872