Daksh Chauhan1, Emeka Anyanwu2, Jacob Goes3, Stephanie A Besser2, Simran Anand4, Ravi Madduri5, Neil Getty6, Sebastian Kelle7, Keigo Kawaji3, Victor Mor-Avi2, Amit R Patel8. 1. University of Chicago, Chicago, IL, United States of America. 2. Department of Medicine, University of Chicago, Chicago, IL, United States of America. 3. Illinois Institute of Technology, Chicago, IL, United States of America. 4. University Pompeu Fabra, Barcelona, Spain. 5. Data Science and Learning Department, Argonne National Laboratory, Lemont, IL, United States of America. 6. Illinois Institute of Technology, Chicago, IL, United States of America; Data Science and Learning Department, Argonne National Laboratory, Lemont, IL, United States of America. 7. Department of Internal Medicine/Cardiology German Heart Center, Berlin, Germany. 8. Department of Medicine, University of Chicago, Chicago, IL, United States of America; Department of Radiology, University of Chicago, Chicago, IL, United States of America. Electronic address: amitpatel@uchicago.edu.
Abstract
BACKGROUND: Artificial intelligence is increasingly utilized to aid in the interpretation of cardiac magnetic resonance (CMR) studies. One of the first steps is the identification of the imaging plane depicted, which can be achieved by both deep learning (DL) and classical machine learning (ML) techniques without user input. We aimed to compare the accuracy of ML and DL for CMR view classification and to identify potential pitfalls during training and testing of the algorithms. METHODS: To train our DL and ML algorithms, we first established datasets by retrospectively selecting 200 CMR cases. The models were trained using two different cohorts (passively and actively curated) and applied data augmentation to enhance training. Once trained, the models were validated on an external dataset, consisting of 20 cases acquired at another center. We then compared accuracy metrics and applied class activation mapping (CAM) to visualize DL model performance. RESULTS: The DL and ML models trained with the passively-curated CMR cohort were 99.1% and 99.3% accurate on the validation set, respectively. However, when tested on the CMR cases with complex anatomy, both models performed poorly. After training and testing our models again on all 200 cases (active cohort), validation on the external dataset resulted in 95% and 90% accuracy, respectively. The CAM analysis depicted heat maps that demonstrated the importance of carefully curating the datasets to be used for training. CONCLUSIONS: Both DL and ML models can accurately classify CMR images, but DL outperformed ML when classifying images with complex heart anatomy.
BACKGROUND: Artificial intelligence is increasingly utilized to aid in the interpretation of cardiac magnetic resonance (CMR) studies. One of the first steps is the identification of the imaging plane depicted, which can be achieved by both deep learning (DL) and classical machine learning (ML) techniques without user input. We aimed to compare the accuracy of ML and DL for CMR view classification and to identify potential pitfalls during training and testing of the algorithms. METHODS: To train our DL and ML algorithms, we first established datasets by retrospectively selecting 200 CMR cases. The models were trained using two different cohorts (passively and actively curated) and applied data augmentation to enhance training. Once trained, the models were validated on an external dataset, consisting of 20 cases acquired at another center. We then compared accuracy metrics and applied class activation mapping (CAM) to visualize DL model performance. RESULTS: The DL and ML models trained with the passively-curated CMR cohort were 99.1% and 99.3% accurate on the validation set, respectively. However, when tested on the CMR cases with complex anatomy, both models performed poorly. After training and testing our models again on all 200 cases (active cohort), validation on the external dataset resulted in 95% and 90% accuracy, respectively. The CAM analysis depicted heat maps that demonstrated the importance of carefully curating the datasets to be used for training. CONCLUSIONS: Both DL and ML models can accurately classify CMR images, but DL outperformed ML when classifying images with complex heart anatomy.
Authors: Mauricio Reyes; Raphael Meier; Sérgio Pereira; Carlos A Silva; Fried-Michael Dahlweid; Hendrik von Tengg-Kobligk; Ronald M Summers; Roland Wiest Journal: Radiol Artif Intell Date: 2020-05-27