Guilherme Moura Cunha1, Kyle A Hasenstab2, Atsushi Higaki3, Kang Wang4, Timo Delgado3, Ryan L Brunsing5, Alexandra Schlein3, Armin Schwartzman6, Albert Hsiao7, Claude B Sirlin3, Katie J Fowler3. 1. Liver Imaging Group, Department of Radiology, University of California San Diego, La Jolla, CA, United States. Electronic address: gcunha@health.ucsd.edu. 2. Liver Imaging Group, Department of Radiology, University of California San Diego, La Jolla, CA, United States; AiDA Laboratory, Department of Radiology, University of California San Diego, La Jolla, CA, United States; Department of Family Medicine and Public Health, University of California San Diego, La Jolla, CA, United States. 3. Liver Imaging Group, Department of Radiology, University of California San Diego, La Jolla, CA, United States. 4. Liver Imaging Group, Department of Radiology, University of California San Diego, La Jolla, CA, United States; AiDA Laboratory, Department of Radiology, University of California San Diego, La Jolla, CA, United States. 5. Radiology, Stanford University, Palo Alto, CA, United States. 6. Department of Family Medicine and Public Health, University of California San Diego, La Jolla, CA, United States. 7. AiDA Laboratory, Department of Radiology, University of California San Diego, La Jolla, CA, United States.
Abstract
PURPOSE: To develop and evaluate the performance of a fully-automated convolutional neural network (CNN)-based algorithm to evaluate hepatobiliary phase (HBP) adequacy of gadoxetate disodium (EOB)-enhanced MRI. Secondarily, we explored the potential of the proposed CNN algorithm to reduce examination length by applying it to EOB-MRI examinations. METHODS: We retrospectively identified EOB-enhanced MRI-HBP series from examinations performed 2011-2018 (internal and external datasets). Our algorithm, comprising a liver segmentation and classification CNN, produces an adequacy score. Two abdominal radiologists independently classified series as adequate or suboptimal. The consensus determination of HBP adequacy was used as ground truth for CNN model training and validation. Reader agreement was evaluated with Cohen's kappa. Performance of the algorithm was assessed by receiver operating characteristics (ROC) analysis and computation of the area under the ROC curve (AUC). Potential examination duration reduction was evaluated descriptively. RESULTS: 1408 HBP series from 484 patients were included. Reader kappa agreement was 0.67 (internal dataset) and 0.80 (external dataset). AUCs were 0.97 (0.96-0.99) for internal and 0.95 (0.92-96) for external and were not significantly different from each other (p = 0.24). 48 % (50/105) examinations could have been shorter by applying the algorithm. CONCLUSION: A proposed CNN-based algorithm achieves higher than 95 % AUC for classifying HBP images as adequate versus suboptimal. The application of this algorithm could potentially shorten examination time and aid radiologists in recognizing technically suboptimal images, avoiding diagnostic pitfalls.
PURPOSE: To develop and evaluate the performance of a fully-automated convolutional neural network (CNN)-based algorithm to evaluate hepatobiliary phase (HBP) adequacy of gadoxetate disodium (EOB)-enhanced MRI. Secondarily, we explored the potential of the proposed CNN algorithm to reduce examination length by applying it to EOB-MRI examinations. METHODS: We retrospectively identified EOB-enhanced MRI-HBP series from examinations performed 2011-2018 (internal and external datasets). Our algorithm, comprising a liver segmentation and classification CNN, produces an adequacy score. Two abdominal radiologists independently classified series as adequate or suboptimal. The consensus determination of HBP adequacy was used as ground truth for CNN model training and validation. Reader agreement was evaluated with Cohen's kappa. Performance of the algorithm was assessed by receiver operating characteristics (ROC) analysis and computation of the area under the ROC curve (AUC). Potential examination duration reduction was evaluated descriptively. RESULTS: 1408 HBP series from 484 patients were included. Reader kappa agreement was 0.67 (internal dataset) and 0.80 (external dataset). AUCs were 0.97 (0.96-0.99) for internal and 0.95 (0.92-96) for external and were not significantly different from each other (p = 0.24). 48 % (50/105) examinations could have been shorter by applying the algorithm. CONCLUSION: A proposed CNN-based algorithm achieves higher than 95 % AUC for classifying HBP images as adequate versus suboptimal. The application of this algorithm could potentially shorten examination time and aid radiologists in recognizing technically suboptimal images, avoiding diagnostic pitfalls.
Authors: Robert M Marks; Andrew Ryan; Elhamy R Heba; An Tang; Tanya J Wolfson; Anthony C Gamst; Claude B Sirlin; Mustafa R Bashir Journal: AJR Am J Roentgenol Date: 2015-03 Impact factor: 3.959
Authors: T Küstner; S Gatidis; A Liebgott; M Schwartz; L Mauch; P Martirosian; H Schmidt; N F Schwenzer; K Nikolaou; F Bamberg; B Yang; F Schick Journal: Magn Reson Imaging Date: 2018-07-21 Impact factor: 2.546
Authors: Ryan L Brunsing; Kathryn J Fowler; Takeshi Yokoo; Guilherme Moura Cunha; Claude B Sirlin; Robert M Marks Journal: Hepatoma Res Date: 2020-09-01